analysis of nike distribution facility’s outbound process flow
TRANSCRIPT
Analysis of Nike Distribution Facility’s Outbound
Submitted in partial fulfilment of the requirements
BACHELORS OF INDUSTRIAL
FACULTY OF ENGINEERING, BUILT ENVIRO
UNIVERSITY OF PRETORIA
Analysis of Nike Distribution Facility’s Outbound
Process Flow
by
Werner Wagenaar
26014964
Submitted in partial fulfilment of the requirements
for the degree of
BACHELORS OF INDUSTRIAL ENGINEERING
in the
FACULTY OF ENGINEERING, BUILT ENVIRONMENT
AND INFORMATION
UNIVERSITY OF PRETORIA
PRETORIA
October 2009
Analysis of Nike Distribution Facility’s Outbound
MENT
Executive Summary
Nike is proud to be associated with the Soccer World Cup 2010 event through the sponsorship
of footwear and kits to various teams and high profile players. Being a leading global brand
Nike is expecting extraordinary demand over the duration of the event. The agility and flexibility
in which Nike respond to these special orders is what will distinguish them f
in this space. The service levels expected by customers for World Cup related orders add
significant pressure to the entire supply chain. The Nike South Africa distribution facility, located
in Meadowdale, is currently not operating at it
flexibility to meet the customer requirements. Barloworld Logistics is contracted to analyse all
distribution processes and develop the capabilities needed to achieve and maintain order
fulfilment throughout the event.
A simulation study was conducted
affected by this event. Along with
vital in the decision making of the build up for the World
not all Key Performance Indicators (
planning had to be reviewed to achieve the desired outcomes
was resistant to change and modification
flexibility.
The change in the resources deployment will save
productivity and utilisation in various
cycle time and order lead time, which in turn will result in increased customer satisfaction.
Increasing the flexibility indicated that more goods could be delivered in the same period of
time, thus increasing overall order fulfilment
Overall the project was a huge succe
yield significant cost reductions.
on unit processing costs and increase throughput capacity.
internally and further studies can be conducted on outbound process flow and planning
processes.
Nike is proud to be associated with the Soccer World Cup 2010 event through the sponsorship
kits to various teams and high profile players. Being a leading global brand
Nike is expecting extraordinary demand over the duration of the event. The agility and flexibility
in which Nike respond to these special orders is what will distinguish them from the competitors
in this space. The service levels expected by customers for World Cup related orders add
significant pressure to the entire supply chain. The Nike South Africa distribution facility, located
in Meadowdale, is currently not operating at its optimal capacity and desired processing
flexibility to meet the customer requirements. Barloworld Logistics is contracted to analyse all
distribution processes and develop the capabilities needed to achieve and maintain order
A simulation study was conducted using Arena® to run different scenarios on process
affected by this event. Along with Microsoft Excel® spreadsheets this simulation proved to be
vital in the decision making of the build up for the World Cup. The results obtained showed that
Key Performance Indicators (KPI’s) where at their optimal levels and resource capacity
planning had to be reviewed to achieve the desired outcomes. It also indicated that the system
modifications on process logic are needed to create the desired
The change in the resources deployment will save Nike labour cost due to
various departments. The system changes will reduce process
rder lead time, which in turn will result in increased customer satisfaction.
Increasing the flexibility indicated that more goods could be delivered in the same period of
overall order fulfilment.
s a huge success and once implemented in February 2010, the project will
Preliminary indications are that the project will yield a reduction
on unit processing costs and increase throughput capacity. The simulation model is re
nternally and further studies can be conducted on outbound process flow and planning
Nike is proud to be associated with the Soccer World Cup 2010 event through the sponsorship
kits to various teams and high profile players. Being a leading global brand,
Nike is expecting extraordinary demand over the duration of the event. The agility and flexibility
m the competitors
in this space. The service levels expected by customers for World Cup related orders add
significant pressure to the entire supply chain. The Nike South Africa distribution facility, located
s optimal capacity and desired processing
flexibility to meet the customer requirements. Barloworld Logistics is contracted to analyse all
distribution processes and develop the capabilities needed to achieve and maintain order
to run different scenarios on processes
spreadsheets this simulation proved to be
ults obtained showed that
s and resource capacity
. It also indicated that the system
d to create the desired
cost due to increased
. The system changes will reduce process
rder lead time, which in turn will result in increased customer satisfaction.
Increasing the flexibility indicated that more goods could be delivered in the same period of
ry 2010, the project will
Preliminary indications are that the project will yield a reduction
The simulation model is re-usable
nternally and further studies can be conducted on outbound process flow and planning
Acknowledgments
TO MY FAMILY AND FRIENDS
ANDRÉ HOUGH
PROF. PAUL KRUGER – PROJECT MENTOR AT
TO MY FAMILY AND FRIENDS – FOR THEIR UNCONDITIONAL LOVE AND SUPPORT
É HOUGH – BARLOWORLD LOGISTICS
PROJECT MENTOR AT THE UNIVERSITY OF PRETORIA
FOR THEIR UNCONDITIONAL LOVE AND SUPPORT
THE UNIVERSITY OF PRETORIA
Table of Contents
1 Introduction ................................
1.1 Background................................
1.1.1 Overview of Outbound Process Flow
1.2 Project Aim ................................
1.3 Project Scope ................................
2 Literature Study................................
2.1 System Analysis Models
2.1.1 Management Decision
2.1.2 Nature of Models ................................
2.2 Basic Approach to Modelling
2.3 Types of Models ................................
2.3.1 Physical Models ................................
2.3.2 Computer Simulation Models
2.3.3 Analytical Models ................................
2.3.4 Analytical versus Simulation Models
2.4 Why Model? ................................
2.5 Simulation Modelling ................................
2.5.1 Simulation Language
2.5.2 High-Level Simulators
2.5.3 Simulation Modelling Tools
3 Conceptual Design of Model
3.1 Structuring the Model ................................
3.2 Conceptual Inputs and Outputs of the Model
i
.........................................................................................................................
................................................................................................
Overview of Outbound Process Flow ................................................................
................................................................................................
................................................................................................
................................................................................................
System Analysis Models ..............................................................................................
Management Decision ..........................................................................................
................................................................................................
Basic Approach to Modelling ........................................................................................
................................................................................................
................................................................................................
Computer Simulation Models ................................................................
................................................................................................
Analytical versus Simulation Models ................................................................
................................................................................................
................................................................................................
Simulation Language ...........................................................................................
Level Simulators ..........................................................................................
Simulation Modelling Tools ................................................................
Conceptual Design of Model ..............................................................................................
................................................................................................
Conceptual Inputs and Outputs of the Model ..............................................................
......................... 1
.................................................. 1
.................................... 3
.................................................. 4
.............................................. 7
................................................... 8
.............................. 8
.......................... 8
.................................. 8
........................ 9
.........................................12
...................................12
...............................................12
.................................13
....................................13
................................................14
...................................15
...........................17
..........................17
..................................................17
..............................22
..................................22
..............................24
4 Data Collection and Analysis
4.1 Input Data and Analysis
4.1.1 Order Information ................................
4.1.2 Transport Time of a Picker per Order
4.1.3 Quantity of a Business Unit Picked
4.1.4 Number to be batched
4.1.5 Distance Conveyor travel in Picking
4.1.6 Process Time Picking
4.1.7 Process Time Packing
4.2 Soft Data ................................
5 Depiction of Simulation Model
5.1 Conceptual Translation
5.1.1 Entities ................................
5.1.2 Attributes ................................
5.1.3 Variables ................................
5.1.4 Resources ................................
5.1.5 Queues ................................
5.1.6 Conveyors ................................
5.1.7 Picker Travel Pattern
6 Simulation Model Construction
6.1 Model Activity Areas ................................
6.1.1 Order Information Generation
6.1.2 Pick Face ................................
6.1.3 Pack Station ................................
6.1.4 Despatch and Output transfer
ii
Data Collection and Analysis..............................................................................................
Input Data and Analysis ..............................................................................................
................................................................................................
Transport Time of a Picker per Order ................................................................
Quantity of a Business Unit Picked ................................................................
Number to be batched .........................................................................................
Distance Conveyor travel in Picking ................................................................
Process Time Picking ..........................................................................................
Process Time Packing .........................................................................................
................................................................................................
Depiction of Simulation Model ............................................................................................
Conceptual Translation ...............................................................................................
................................................................................................
................................................................................................
................................................................................................
................................................................................................
................................................................................................
................................................................................................
Picker Travel Pattern ...........................................................................................
Simulation Model Construction ...........................................................................................
................................................................................................
Order Information Generation ................................................................
................................................................................................
................................................................................................
and Output transfer ................................................................
..............................27
..............................27
.................................27
...................................28
......................................30
.........................30
.....................................31
..........................34
.........................35
.....................................................36
............................37
...............................37
.................................................37
.............................................37
..............................................39
...........................................40
................................................41
...........................................42
...........................42
...........................44
...................................44
..............................................44
.............................................47
........................................50
..............................................52
6.2 Verification and Validation of the Model
6.2.1 Verification ................................
6.2.2 Validation ................................
6.3 Scenario Development ................................
6.3.1 Increased growth in forthcoming years
6.3.2 Change in the number of resources
6.3.3 Change the number of orders released from the pick pool
6.4 Possible follow-up Scenarios
6.4.1 Business unit locations
6.4.2 Conveyor Speed ................................
7 Simulation Results and Recommendations
7.1 Increase growth in forthcoming years
7.1.1 Scenario 1: A ................................
7.1.2 Scenario 1: B ................................
7.1.3 Scenario 1: C ................................
7.2 Change in resource capacity
7.2.1 Scenario 2: A ................................
7.2.2 Scenario 2: B ................................
7.2.3 Scenario 2: C ................................
7.3 Change the batch size of orders send from the pick pool
8 Conclusion ................................
9 References ................................
10 Appendices ................................
10.1 Input Data Analysis ................................
10.2 Probability Distributions
iii
Verification and Validation of the Model ................................................................
................................................................................................
................................................................................................
................................................................................................
Increased growth in forthcoming years................................................................
Change in the number of resources ................................................................
Change the number of orders released from the pick pool ................................
up Scenarios ................................................................
Business unit locations ........................................................................................
................................................................................................
Simulation Results and Recommendations ................................................................
Increase growth in forthcoming years................................................................
................................................................................................
................................................................................................
................................................................................................
n resource capacity .......................................................................................
................................................................................................
................................................................................................
................................................................................................
he batch size of orders send from the pick pool ................................
.........................................................................................................................
........................................................................................................................
................................................................................................
................................................................................................
Probability Distributions ..............................................................................................
......................................53
...........................................53
.............................................54
................................55
.................................55
.....................................56
...................................56
......................................................57
........................57
..................................57
........................................58
..........................................58
.......................................58
.......................................59
.......................................61
.......................63
.......................................63
.......................................64
.......................................66
............................................68
.........................70
........................71
....................................................73
.....................................73
..............................77
10.2.1 Picking Process Time for Business Unit “Other” Distribution
10.2.2 Picking Process Time for Business Unit “Footwear” Distribution
10.2.3 Packing Process Time Distribution
10.3 Simulation Results ................................
10.3.1 Queue average waiting time per entity
10.3.2 Number of entities entering a process
10.4 The Simulation Parts ................................
10.4.1 Order information generation phase
10.4.2 Pick Face ................................
10.4.3 Pack Station ................................
10.4.4 Despatch and Output transfer phase
iv
Picking Process Time for Business Unit “Other” Distribution ................................
Picking Process Time for Business Unit “Footwear” Distribution ..........................
Packing Process Time Distribution ................................................................
................................................................................................
Queue average waiting time per entity ................................................................
Number of entities entering a process ................................................................
................................................................................................
Order information generation phase ................................................................
................................................................................................
................................................................................................
tch and Output transfer phase ................................................................
................................77
..........................78
.......................................79
......................................80
.................................80
..................................81
...................................82
.....................................82
.............................................83
........................................84
...................................85
List of Figures
Figure 1: Process Flow Diagram
Figure 2: Outbound Process Flow
Figure 3: Scope Boundary ................................
Figure 4: Basic Approach to Modelling
Figure 5: Areas influenced by modelling
Figure 6: Illustrating the mapping of simulation between the real world and models
Figure 7: Major Phases in Simulation S
Figure 8: Flow chart of an Arrival
Figure 9: Walking time from one block to next
Figure 10: Illustration of the split of an Order
Figure 11: Illustration of Conveyor Movement on Picking Levels
Figure 12: Level Layout ................................
Figure 13: Data input to Simulation Study
Figure 14: Picker Travel Pattern ................................
Figure 15: Read Simulation Model Variables
Figure 16: Duplication of Entity to the Number of Orders
Figure 17: Assigning and Reading Attributes to Entity
Figure 18: Assigning Orders to different l
Figure 19: Picking and Conveying Processes
Figure 20: Decide on packing station
Figure 21: Packing Process ................................
Figure 22: Delay, Batching and Output transfer processes
Figure 23: Despatching of orders
Figure 24: Resource utilisation of base model with 8 pickers
Figure 25: Resource utilisation of three year projection
Figure 26: Resource utilisation of six year projection
Figure 27: Resource utilisation with increase of three pickers
Figure 28: Resource utilisation with increase of seven pickers
Figure 29: Resource utilisation with decrease of three packing stations
v
Figure 1: Process Flow Diagram ................................................................................................
Figure 2: Outbound Process Flow ..............................................................................................
................................................................................................
Figure 4: Basic Approach to Modelling ................................................................
Figure 5: Areas influenced by modelling................................................................
Figure 6: Illustrating the mapping of simulation between the real world and models
Figure 7: Major Phases in Simulation Study ................................................................
Figure 8: Flow chart of an Arrival...............................................................................................
Figure 9: Walking time from one block to next ................................................................
Figure 10: Illustration of the split of an Order ................................................................
Figure 11: Illustration of Conveyor Movement on Picking Levels ................................
................................................................................................
Figure 13: Data input to Simulation Study ................................................................
................................................................................................
Figure 15: Read Simulation Model Variables ................................................................
Figure 16: Duplication of Entity to the Number of Orders ..........................................................
Figure 17: Assigning and Reading Attributes to Entity ...............................................................
Figure 18: Assigning Orders to different levels for picking .........................................................
Figure 19: Picking and Conveying Processes ................................................................
Figure 20: Decide on packing station ........................................................................................
................................................................................................
Figure 22: Delay, Batching and Output transfer processes .......................................................
Figure 23: Despatching of orders ..............................................................................................
Resource utilisation of base model with 8 pickers ................................
Figure 25: Resource utilisation of three year projection .............................................................
Figure 26: Resource utilisation of six year projection ................................................................
Figure 27: Resource utilisation with increase of three pickers ................................
esource utilisation with increase of seven pickers ................................
Figure 29: Resource utilisation with decrease of three packing stations ................................
................................ 3
.............................. 4
......................................... 7
......................................................11
....................................................15
Figure 6: Illustrating the mapping of simulation between the real world and models ..................16
..............................................23
...............................25
...........................................29
.............................................31
...............................................32
.............................................33
.................................................36
................................43
............................................45
..........................46
...............................47
.........................48
...........................................49
........................50
......................................51
.......................52
..............................53
....................................................59
.............................60
................................62
...................................................64
..................................................65
....................................67
Figure 30: Decrease in orders send from pick pool results
Figure 31: Resource utilisation with decrease in orders send from pick pool
List of Tables
Table 1: Distance Conveyor travel from each block
Table 2: Processing Times for Picking in sec/unit
Table 3: Processing Times for Packing
Table 4: List of Attributes ................................
Table 5: List of Variables ................................
Table 6: List of Resources ................................
Table 7: List of Queues ................................
Table 8: Conveyors ................................
Table 9: Base model results ................................
Table 10: Three year projection in quantity of units ordered results
Table 11: Six year projection in quantity of units ordered results
Table 12: Increase picker resource capacity by three results
Table 13: Increase picker resource capacity by seven results
Table 14: Decrease packer resource capacity by three results
Table 15: Picker Transport Time Table fo
Table 16: Picker Transport Time Table for Level B, C, D
Table 17: Extraction from Total Transport Time for Each Level in Picking
Table 18: Extraction from Quantity and Business Units per Order on each Level
Table 19: Extraction from Number to be batched on a Level
Table 20: Extraction Distance the Conveyor travel on a Level
vi
Figure 30: Decrease in orders send from pick pool results ........................................................
Figure 31: Resource utilisation with decrease in orders send from pick pool .............................
Table 1: Distance Conveyor travel from each block ................................................................
Picking in sec/unit ................................................................
Table 3: Processing Times for Packing ................................................................
................................................................................................
................................................................................................
................................................................................................
................................................................................................
................................................................................................
................................................................................................
n in quantity of units ordered results................................
Table 11: Six year projection in quantity of units ordered results ................................
Table 12: Increase picker resource capacity by three results ................................
Table 13: Increase picker resource capacity by seven results ................................
Table 14: Decrease packer resource capacity by three results ................................
Table 15: Picker Transport Time Table for Level A ................................................................
Table 16: Picker Transport Time Table for Level B, C, D ..........................................................
Table 17: Extraction from Total Transport Time for Each Level in Picking ................................
Table 18: Extraction from Quantity and Business Units per Order on each Level
Table 19: Extraction from Number to be batched on a Level ................................
Table 20: Extraction Distance the Conveyor travel on a Level ................................
........................68
.............................69
..................................34
.....................................35
.....................................................35
..........................................38
...........................................39
........................................40
.............................................41
...................................................42
......................................58
...........................................60
...............................................61
....................................................63
...................................................65
.................................................66
...................................73
..........................74
.................................75
Table 18: Extraction from Quantity and Business Units per Order on each Level ......................75
.....................................................76
...................................................76
1 Introduction
1.1 Background
Nike, Inc. is a global company operating in 4 regions and 6 continents and with global revenue
exceeding US $14 billion in 2006. Nike’s global headquarters are located in Beaverton in the
USA. South Africa is part of the EMEA (Europe, Middle
headquartered in Hilversum, the Netherlands.
Originally managed by a distributor, Nike decided to manage marketing and distribution in South
Africa through its own organisation which was implemented in 1995. The office is loc
Midrand, Johannesburg with a satellite sales office in Cape Town. The total organisation
employs 120 people with a turnover of approximately
classified in 3 categories:
� Footwear
� Categories such as Running, Basketb
� Apparel
� Products like Shirts, Tank Tops, Jerseys, Track Suits, Windbreakers, Sports
Jackets and Caps
� Equipment
� Items such as Socks, Bags, Back Packs, Gloves, Timing and Balls
Nike enjoys major market shares
Adidas, Ellesse and Puma in this space.
1
Nike, Inc. is a global company operating in 4 regions and 6 continents and with global revenue
exceeding US $14 billion in 2006. Nike’s global headquarters are located in Beaverton in the
USA. South Africa is part of the EMEA (Europe, Middle-East and Africa) region which is
headquartered in Hilversum, the Netherlands.
Originally managed by a distributor, Nike decided to manage marketing and distribution in South
Africa through its own organisation which was implemented in 1995. The office is loc
, Johannesburg with a satellite sales office in Cape Town. The total organisation
with a turnover of approximately $100m. Nike’s product
Categories such as Running, Basketball, Football, Cross-Training and Tennis
Products like Shirts, Tank Tops, Jerseys, Track Suits, Windbreakers, Sports
Jackets and Caps
Items such as Socks, Bags, Back Packs, Gloves, Timing and Balls
shares across all 3 business units’. They compete
in this space.
Nike, Inc. is a global company operating in 4 regions and 6 continents and with global revenue
exceeding US $14 billion in 2006. Nike’s global headquarters are located in Beaverton in the
t and Africa) region which is
Originally managed by a distributor, Nike decided to manage marketing and distribution in South
Africa through its own organisation which was implemented in 1995. The office is located in
, Johannesburg with a satellite sales office in Cape Town. The total organisation
product range can be
Training and Tennis
Products like Shirts, Tank Tops, Jerseys, Track Suits, Windbreakers, Sports
Items such as Socks, Bags, Back Packs, Gloves, Timing and Balls
They compete with Reebok,
In South Africa, product reaches customers through the following distribution channels:
• National and strategic accounts of multi
• Franchised Nike only stores
• Nike owned Factory outlets selling close
inventory
Nike’s distribution facility is loca
Logistics was contracted by Nike to investigate the local di
expected increase of sales during the 2010 Soccer World Cup in South Africa. I
that at least 50,000 units per day must be distributed through the distributions facility, to meet
the expected demand. The distr
30,000 units per day, which is far less than the projected need. Nike is a major sponsor of the
Soccer World Cup, which places an important responsibility on the distribution facility to run
its optimal level, thereby insuring a constant av
reduced share in the market to its various competitors. Nike’s ability to respond to a rapid
increase and decrease in demand is an important factor in its
competitors and retain market share.
The focus of Barloworld Logistics is on the distribution facility and all the processes it entails.
Thus an overview is needed of the facility and the recommendations to ensure that the
will be achieved for the Soccer World Cup 2010
2
In South Africa, product reaches customers through the following distribution channels:
National and strategic accounts of multi-brand stores
stores
Nike owned Factory outlets selling close-out (also called aged) inventory and excess
Nike’s distribution facility is located in Meadowdale, Johannesburg. During 2009
by Nike to investigate the local distribution facility in view of an
expected increase of sales during the 2010 Soccer World Cup in South Africa. I
000 units per day must be distributed through the distributions facility, to meet
the expected demand. The distribution facility is currently running at a level of approximately
000 units per day, which is far less than the projected need. Nike is a major sponsor of the
Soccer World Cup, which places an important responsibility on the distribution facility to run
its optimal level, thereby insuring a constant availability of products, failing
reduced share in the market to its various competitors. Nike’s ability to respond to a rapid
increase and decrease in demand is an important factor in its ability to compete with the main
competitors and retain market share.
The focus of Barloworld Logistics is on the distribution facility and all the processes it entails.
Thus an overview is needed of the facility and the recommendations to ensure that the
will be achieved for the Soccer World Cup 2010 season.
In South Africa, product reaches customers through the following distribution channels:
inventory and excess
During 2009 Barloworld
stribution facility in view of an
expected increase of sales during the 2010 Soccer World Cup in South Africa. It is estimated
000 units per day must be distributed through the distributions facility, to meet
at a level of approximately
000 units per day, which is far less than the projected need. Nike is a major sponsor of the
Soccer World Cup, which places an important responsibility on the distribution facility to run at
might lead to a
reduced share in the market to its various competitors. Nike’s ability to respond to a rapid
ability to compete with the main
The focus of Barloworld Logistics is on the distribution facility and all the processes it entails.
Thus an overview is needed of the facility and the recommendations to ensure that the demand
1.1.1 Overview of Outbound
In the distribution facility there are mainly two types of process flows. The first is the Inbound
Process Flow (Figure 1); this process flow takes into account all the process
and resources in receiving of goods from the suppliers. The departments running the Inbound
Process Flow are the Goods Rec
and the High Bay Department (the storing of the goods in bulk
The second process flow is the Outbound Process Flow (
account all the processes, flow of material and resources in despatching specific orders to
customers. The departments running the Outbound Process Flow are
Value Added Service (VAS), Packing department and the Despatching department.
Figure 1: Process Flow Diagram
On a higher level of detail the Outbound Process Flow consists of an order that comes from a
customer. This order contains all the necessary detail from the products to be purchased to the
delivery date of the products. After the order is complete it goes to the Pick Pool
From the Pick Pool it is taken to the Wave planning, where the planning occurs of when each
order must be picked in order to reach the specified delivery date.
constraints into account from the
delivery plan) which is the less urgent orders.
3
Overview of Outbound Process Flow
there are mainly two types of process flows. The first is the Inbound
; this process flow takes into account all the processes
and resources in receiving of goods from the suppliers. The departments running the Inbound
he Goods Receiving department (the receiving of goods from the supplier)
and the High Bay Department (the storing of the goods in bulk storage locations
The second process flow is the Outbound Process Flow (Figure 1). This process flow takes into
account all the processes, flow of material and resources in despatching specific orders to
customers. The departments running the Outbound Process Flow are the Fine P
Value Added Service (VAS), Packing department and the Despatching department.
On a higher level of detail the Outbound Process Flow consists of an order that comes from a
customer. This order contains all the necessary detail from the products to be purchased to the
of the products. After the order is complete it goes to the Pick Pool
From the Pick Pool it is taken to the Wave planning, where the planning occurs of when each
picked in order to reach the specified delivery date. This planning takes di
from the, expedite, which is the most urgent to the Non IDP (Integrated
urgent orders. These orders get released by Wave planning and
there are mainly two types of process flows. The first is the Inbound
es, flow of material
and resources in receiving of goods from the suppliers. The departments running the Inbound
of goods from the supplier)
storage locations).
his process flow takes into
account all the processes, flow of material and resources in despatching specific orders to
Pick department,
Value Added Service (VAS), Packing department and the Despatching department.
On a higher level of detail the Outbound Process Flow consists of an order that comes from a
customer. This order contains all the necessary detail from the products to be purchased to the
of the products. After the order is complete it goes to the Pick Pool (Figure 2).
From the Pick Pool it is taken to the Wave planning, where the planning occurs of when each
planning takes different
expedite, which is the most urgent to the Non IDP (Integrated
These orders get released by Wave planning and
Orders
Pick Pool
Wave
Planning
the picking of the goods takes place. After the goods have been picked it is sent to th
department. At the Packing Department the goods are packed into boxes and sent to the
Despatch department where it gets despatched to the
Figure 2: Outbound Process Flow
1.2 Project Aim
The main objective of this project is to provide Barloworld Logistics with a model to assess if the
outbound process flow of the distribution facility would meet the output and flexibility
requirements over the Soccer W
achieved for the increased sales o
any, which might take place at the distribution facility to run at optimum throughput capacity.
The facility was designed to achieve a theoretical output of 50,000
due to the expected high demand
be as close as possible to the theoretical output per day. The process analysis will include
demand variability, seasonality and
important and peak demand period.
4
Release
Picking
Packing
Despaching
the goods takes place. After the goods have been picked it is sent to th
At the Packing Department the goods are packed into boxes and sent to the
Despatch department where it gets despatched to the specific customer (Figure
: Outbound Process Flow
of this project is to provide Barloworld Logistics with a model to assess if the
of the distribution facility would meet the output and flexibility
World Cup period. It will also assess if the demand levels can be
achieved for the increased sales over the Soccer World Cup season, as well as any changes, if
any, which might take place at the distribution facility to run at optimum throughput capacity.
as designed to achieve a theoretical output of 50,000 units per day. It is essential
due to the expected high demand of the Soccer World Cup that the actual output per day must
be as close as possible to the theoretical output per day. The process analysis will include
demand variability, seasonality and growth over the next few months, which
demand period.
Despaching
the goods takes place. After the goods have been picked it is sent to the Packing
At the Packing Department the goods are packed into boxes and sent to the
Figure 2).
of this project is to provide Barloworld Logistics with a model to assess if the
of the distribution facility would meet the output and flexibility
. It will also assess if the demand levels can be
, as well as any changes, if
any, which might take place at the distribution facility to run at optimum throughput capacity.
units per day. It is essential
of the Soccer World Cup that the actual output per day must
be as close as possible to the theoretical output per day. The process analysis will include
growth over the next few months, which are Nike’s most
An effective response to the changes in customer demand is the decisive factor. The demand
will increase during the build-up and the occurrence of the Soccer World Cup 2010. With
continuous changes comes a need for intuitive and fle
great importance for the Soccer World Cup. If a team goes through to the n
Soccer World Cup, it is expected that the demand for
They will be ordered today and need to be
orders are of higher priority and the flexibility of
met.
The simulation model will be used in studying accurate information and recommendati
management pertaining to the following decision making parameters:
• Highlight the possible bottlenecks under a certain set of conditions
• Flexibility of the system under changes in demand
• Material flow in the distribution facility
• Number of staff needed at each station
• Increase throughput, as close as possible to the theoretical maximum per day
• Calculate the buffer sizes, if needed, at each station
• Increase resource utilisation
• Inter relationship between processes
5
An effective response to the changes in customer demand is the decisive factor. The demand
up and the occurrence of the Soccer World Cup 2010. With
continuous changes comes a need for intuitive and flexible control. A highly flexible
great importance for the Soccer World Cup. If a team goes through to the n
it is expected that the demand for Nike products of that team will increase.
ay and need to be despatched the next day, this indicates that some
and the flexibility of the system will indicate if the demand can be
The simulation model will be used in studying accurate information and recommendati
management pertaining to the following decision making parameters:
Highlight the possible bottlenecks under a certain set of conditions
Flexibility of the system under changes in demand
Material flow in the distribution facility
ded at each station
Increase throughput, as close as possible to the theoretical maximum per day
Calculate the buffer sizes, if needed, at each station
utilisation
Inter relationship between processes
An effective response to the changes in customer demand is the decisive factor. The demand
up and the occurrence of the Soccer World Cup 2010. With
flexible system is of
great importance for the Soccer World Cup. If a team goes through to the next round in the
that team will increase.
the next day, this indicates that some
if the demand can be
The simulation model will be used in studying accurate information and recommendations to
Increase throughput, as close as possible to the theoretical maximum per day
The essence of the parameter will entail the following processes:
� Optimising the processes (
� Material handling, the flow of material through the facility
� Line balancing
� Queuing analysis.
The ultimate goal of the study is to provide Barloworld Logistics wi
the distribution facility’s readiness to cope with the peak demand for the Soccer World Cup. The
study will also indicate what recommendations and changes must be made in order to achieve
that goal.
6
will entail the following processes:
Optimising the processes (E.g. Throughput optimisation)
Material handling, the flow of material through the facility
The ultimate goal of the study is to provide Barloworld Logistics with a model, which will indicate
the distribution facility’s readiness to cope with the peak demand for the Soccer World Cup. The
study will also indicate what recommendations and changes must be made in order to achieve
th a model, which will indicate
the distribution facility’s readiness to cope with the peak demand for the Soccer World Cup. The
study will also indicate what recommendations and changes must be made in order to achieve
1.3 Project Scope
All operation activities regarding the
2). The project is seen as a tactical model for the distribution facility. It will look at the facility and
its outbound operations on a low
its operations. The return station is seen as an operation on its own and the
in the outbound flow of goods and thus will be not be included in the model.
Historical data are used and not the methods of how those data where collected. In essence all
outbound processes are included, from
goods (Figure 2). Figure 3 more clearly illustrates the border of the scope
Figure 3: Scope Boundary
7
ation activities regarding the Outbound Process Flow are included in this project
a tactical model for the distribution facility. It will look at the facility and
low level, taking into account smaller details of every station and
its operations. The return station is seen as an operation on its own and therefore does not fall
flow of goods and thus will be not be included in the model.
Historical data are used and not the methods of how those data where collected. In essence all
are included, from where the orders are made until the despatching of
more clearly illustrates the border of the scope.
are included in this project (Figure
a tactical model for the distribution facility. It will look at the facility and
level, taking into account smaller details of every station and
refore does not fall
Historical data are used and not the methods of how those data where collected. In essence all
until the despatching of
2 Literature Study
2.1 System Analysis Models
2.1.1 Management Decision
It is essential for managers to base their decisions and the extent to which their
made, on the time horizon. The event of the Soccer World Cup takes place in a few months time
and in order to satisfy the demand it is important to make informed decisions. According to
Buzacott & Shanthikumar (1993:5) it is useful to disting
operational decisions, which are defined as follows:
� Strategic – Strategic decisions are those with long time horizon (more than a year or
two) and involve major commitments of the organisation’s resources. These decisions
must be in alignment with the corporate strategy.
� Tactical – Tactical decisions have an interme
months or a lesser scope. These tactical decisions, in turn, become the operating
constraints under which operational planning and control decisions are made.
� Operational – Operational decisions are typically those ma
2.1.2 Nature of Models
According to Buzacott & Shanthikumar (1993:8) a model is based on a representation of the
operation of the system. Buzacott & Shanthikumar (1993:8) states that, the model is ei
mathematical formula or a computer programme that when supplied with the numerical values
of various parameters will make a numerical prediction of the system performance measures.
By changing the values of a parameter the user can gain insight into
performance. Models are intended to support management decisions about the system.
8
System Analysis Models
Management Decision
It is essential for managers to base their decisions and the extent to which their
made, on the time horizon. The event of the Soccer World Cup takes place in a few months time
and in order to satisfy the demand it is important to make informed decisions. According to
Buzacott & Shanthikumar (1993:5) it is useful to distinguish between strategic, tactical and
operational decisions, which are defined as follows:
Strategic decisions are those with long time horizon (more than a year or
two) and involve major commitments of the organisation’s resources. These decisions
must be in alignment with the corporate strategy.
Tactical decisions have an intermediate time horizon of between 3 and 18
months or a lesser scope. These tactical decisions, in turn, become the operating
constraints under which operational planning and control decisions are made.
Operational decisions are typically those made every day or every week.
(Chase, Jacobs & Aquilano 2006:9
According to Buzacott & Shanthikumar (1993:8) a model is based on a representation of the
operation of the system. Buzacott & Shanthikumar (1993:8) states that, the model is ei
mathematical formula or a computer programme that when supplied with the numerical values
of various parameters will make a numerical prediction of the system performance measures.
By changing the values of a parameter the user can gain insight into
performance. Models are intended to support management decisions about the system.
It is essential for managers to base their decisions and the extent to which their decisions are
made, on the time horizon. The event of the Soccer World Cup takes place in a few months time
and in order to satisfy the demand it is important to make informed decisions. According to
uish between strategic, tactical and
Strategic decisions are those with long time horizon (more than a year or
two) and involve major commitments of the organisation’s resources. These decisions
diate time horizon of between 3 and 18
months or a lesser scope. These tactical decisions, in turn, become the operating
constraints under which operational planning and control decisions are made.
de every day or every week.
Chase, Jacobs & Aquilano 2006:9)
According to Buzacott & Shanthikumar (1993:8) a model is based on a representation of the
operation of the system. Buzacott & Shanthikumar (1993:8) states that, the model is either a
mathematical formula or a computer programme that when supplied with the numerical values
of various parameters will make a numerical prediction of the system performance measures.
its influence on
performance. Models are intended to support management decisions about the system.
2.2 Basic Approach to Modelling
The process of modelling involves the following steps:
� Identify the issues to be addressed
The needs of management must be e
The decisions that the model is required to support have to be identified.
� Learn about the system
Buzacott & Shanthikumar
the machines, material handling and the data collection and control systems. Determine
the characteristics of the job and the target volumes, quality and cost
review any existing models to identify their capabilities, shortcomings and the level of
familiarity of managers with the use of
design or operation.
� Choose a modelling approach
“Various types of modelling
models through computer simulations
by the time and cost budget
level of user interface should also be considered when selecting the
� Develop and test the model
During this step, data on the parameters of the model should be acquired. Decisions on
the level of detail, accuracy and the method of collecting data should be made, since it
will influence the correctness of the system, Buzacott & Shanthikumar (1993:8).
9
Modelling
involves the following steps:
Identify the issues to be addressed
The needs of management must be established before clarifying the major concerns.
The decisions that the model is required to support have to be identified.
Learn about the system
Buzacott & Shanthikumar (1993:8) said, “Identify the elements of the system, such as
handling and the data collection and control systems. Determine
the characteristics of the job and the target volumes, quality and cost
review any existing models to identify their capabilities, shortcomings and the level of
f managers with the use of modelling approaches to help in the system
approach
modelling approaches can be used, ranging from formal mathematical
models through computer simulations. The choice of modelling approach is determined
by the time and cost budget” said by said by Buzacott & Shanthikumar (1993:8)
level of user interface should also be considered when selecting the modelling
Develop and test the model
on the parameters of the model should be acquired. Decisions on
the level of detail, accuracy and the method of collecting data should be made, since it
e the correctness of the system, Buzacott & Shanthikumar (1993:8).
stablished before clarifying the major concerns.
The decisions that the model is required to support have to be identified.
the elements of the system, such as
handling and the data collection and control systems. Determine
the characteristics of the job and the target volumes, quality and cost”. It is useful to
review any existing models to identify their capabilities, shortcomings and the level of
approaches to help in the system
approaches can be used, ranging from formal mathematical
approach is determined
said by said by Buzacott & Shanthikumar (1993:8). The
modelling approach.
on the parameters of the model should be acquired. Decisions on
the level of detail, accuracy and the method of collecting data should be made, since it
e the correctness of the system, Buzacott & Shanthikumar (1993:8).
� Verify and validate the model
The model should be checked to see that it is a reasonably correct representation of
reality; a base case should be established. This includes verification and validation.
• “Verification is to ensure the model results are correct derived from the
assumptions made in developing the model
(1993:8).
• “Validation is the process of ensuring that the assumptions of the model and the
results are an accurate representation of the real system
Shanthikumar (1993:8).
� Develop a model interface for the user
Buzacott & Shanthikumar (1993:8) said “
decisions it has to be provided with some interface so that it can be used by
management”. This requires the modeller to eit
support system or to present the model and its implications in a way that management
can understand. The user has to be convinced that the model is adequate for the
decision that the model is required to support.
� Experiment with the model
“This requires exploring the impact of changes in model parameters and developing
understanding of the factors influencing performance of the system so that the manager
can be confident in the decisions made using the model
Shanthikumar (1993:8).
10
the model
The model should be checked to see that it is a reasonably correct representation of
reality; a base case should be established. This includes verification and validation.
Verification is to ensure the model results are correct derived from the
assumptions made in developing the model” said by Buzacott & Shanthikumar
Validation is the process of ensuring that the assumptions of the model and the
results are an accurate representation of the real system” said by Buzacott &
r (1993:8).
Develop a model interface for the user
Buzacott & Shanthikumar (1993:8) said “If the model is to be of value in making
decisions it has to be provided with some interface so that it can be used by
. This requires the modeller to either incorporate the model into a decision
support system or to present the model and its implications in a way that management
can understand. The user has to be convinced that the model is adequate for the
decision that the model is required to support.
Experiment with the model
This requires exploring the impact of changes in model parameters and developing
understanding of the factors influencing performance of the system so that the manager
can be confident in the decisions made using the model” said
The model should be checked to see that it is a reasonably correct representation of
reality; a base case should be established. This includes verification and validation.
Verification is to ensure the model results are correct derived from the
” said by Buzacott & Shanthikumar
Validation is the process of ensuring that the assumptions of the model and the
” said by Buzacott &
If the model is to be of value in making
decisions it has to be provided with some interface so that it can be used by
her incorporate the model into a decision
support system or to present the model and its implications in a way that management
can understand. The user has to be convinced that the model is adequate for the
This requires exploring the impact of changes in model parameters and developing
understanding of the factors influencing performance of the system so that the manager
” said by Buzacott &
� Present the Results
Using the model, the manager should implement the recommended course of action. It
is important that the users are informed about the accuracy of the results which will be
generated by the model
Shanthikumar (1993:8).
Figure 4: Basic Approach to Modelling
11
Using the model, the manager should implement the recommended course of action. It
is important that the users are informed about the accuracy of the results which will be
in order to make correct and informed decisions, Buzacott &
Modelling
Using the model, the manager should implement the recommended course of action. It
is important that the users are informed about the accuracy of the results which will be
correct and informed decisions, Buzacott &
2.3 Types of Models
For discrete part manufacturing there are three types of models in common use:
2.3.1 Physical Models
(Buzacott & Shanthikumar 1993:11) said, “
another physical system, in which parts move from one machine to another and the machines
perform processing operations on the parts. The major difference
the model uses a different dimensional scale, so a large system will occupy a large amount
space. Physical models are excellent as a means of educating production management and
workers about the control of the system, but the
run behaviour of the system as it is difficult to represent the statistical properties of events such
as machine failures”.
2.3.2 Computer Simulation Models
Kelton, Sadowski & Sturrock (2007:5) said that, “
studying a wide variety of models of real
designed to imitate the system’s operations or characteristics, often over time. From a practical
viewpoint, simulation is the process of designing and creating a computerised model of a real or
proposed system for the purpose of conducting numerical experiments to give us a better
understanding of the behaviour of that system for a given set of conditions
can become quite complicated, if needed to represent the system faithfully and you can
simulation analysis.
“The probabilistic nature of many events, such as machine failure, can be represented by
sampling from a distribution representing the patte
represent the typical behaviour of the system it is necessary to run the simulation model for a
sufficient long time, so that all events can occur a reasonable number of times
Buzacott & Shanthikumar (1993:1
graphic display to demonstrate the movement of jobs and material handling
12
For discrete part manufacturing there are three types of models in common use:
(Buzacott & Shanthikumar 1993:11) said, “Physical models represent the real system by
another physical system, in which parts move from one machine to another and the machines
perform processing operations on the parts. The major difference from the real system is that
the model uses a different dimensional scale, so a large system will occupy a large amount
space. Physical models are excellent as a means of educating production management and
workers about the control of the system, but they do not lend themselves to assessing the long
run behaviour of the system as it is difficult to represent the statistical properties of events such
Computer Simulation Models
Kelton, Sadowski & Sturrock (2007:5) said that, “Computer simulation refers to methods for
studying a wide variety of models of real-world systems by numerical evaluation using software
designed to imitate the system’s operations or characteristics, often over time. From a practical
process of designing and creating a computerised model of a real or
proposed system for the purpose of conducting numerical experiments to give us a better
understanding of the behaviour of that system for a given set of conditions”. Simulation models
, if needed to represent the system faithfully and you can
The probabilistic nature of many events, such as machine failure, can be represented by
sampling from a distribution representing the pattern of occurrence of the event. Thus to
represent the typical behaviour of the system it is necessary to run the simulation model for a
sufficient long time, so that all events can occur a reasonable number of times
Buzacott & Shanthikumar (1993:11). Simulation models can be provided with an interactive
graphic display to demonstrate the movement of jobs and material handling devices. This can
For discrete part manufacturing there are three types of models in common use:
Physical models represent the real system by
another physical system, in which parts move from one machine to another and the machines
from the real system is that
the model uses a different dimensional scale, so a large system will occupy a large amount
space. Physical models are excellent as a means of educating production management and
y do not lend themselves to assessing the long
run behaviour of the system as it is difficult to represent the statistical properties of events such
simulation refers to methods for
world systems by numerical evaluation using software
designed to imitate the system’s operations or characteristics, often over time. From a practical
process of designing and creating a computerised model of a real or
proposed system for the purpose of conducting numerical experiments to give us a better
. Simulation models
, if needed to represent the system faithfully and you can still do a
The probabilistic nature of many events, such as machine failure, can be represented by
rn of occurrence of the event. Thus to
represent the typical behaviour of the system it is necessary to run the simulation model for a
sufficient long time, so that all events can occur a reasonable number of times”, said by
. Simulation models can be provided with an interactive
devices. This can
be of great value in communicating the assumptions of the model to the managers (Buzacott &
Shanthikumar 1993:11).
2.3.3 Analytical Models
Analytical models describe the system using mathematical or symbolic relationships. If the
model is simple, traditional mathematical tools like queuing theory, differential
or linear programming can be used (K
Shanthikumar 1993:12 said that, “
algorithm or computational procedure by which the performance measures of the system can be
calculated. Analytical models can also be used to demonstrate properties of various operating
rules and control strategies”.
2.3.4 Analytical versus Simulation Models
Simulation models can be developed in several different ways. One is to try and include all the
required detail of the system from the beginning. This tends to result in long development and
debugging time, and usually makes the resultant code unreadable to all but the developer.
Another approach is to develop models of the system components and after each component
model has been debugged, tested and validated, link the components two at a time, test and
validate, then three at a time, test and validate and so on. Sometimes there are already existing
simulation models of various sub
overall system model but often it is possible to devise ways of summarising the performance of
the sub-system model and replacing it by a simple sub
approach is called the micro-macro approach and is useful
systems.
Simulation models are usually written in a simulation language such as GPSS, SLAM, SIMAN,
SIMULA or SMALLTALK and only experts in the simulation language can read the code.
13
be of great value in communicating the assumptions of the model to the managers (Buzacott &
Analytical models describe the system using mathematical or symbolic relationships. If the
model is simple, traditional mathematical tools like queuing theory, differential-equation methods
or linear programming can be used (Kelton, Sadowski & Sturrock 2007:5).
Shanthikumar 1993:12 said that, “These are then used to derive a formula or to define an
algorithm or computational procedure by which the performance measures of the system can be
s can also be used to demonstrate properties of various operating
Analytical versus Simulation Models
Simulation models can be developed in several different ways. One is to try and include all the
tem from the beginning. This tends to result in long development and
debugging time, and usually makes the resultant code unreadable to all but the developer.
Another approach is to develop models of the system components and after each component
been debugged, tested and validated, link the components two at a time, test and
validate, then three at a time, test and validate and so on. Sometimes there are already existing
simulation models of various sub-systems. The problem is that these can be too detailed for the
overall system model but often it is possible to devise ways of summarising the performance of
system model and replacing it by a simple sub-system component model. This
macro approach and is useful in developing models of very large
Simulation models are usually written in a simulation language such as GPSS, SLAM, SIMAN,
SIMULA or SMALLTALK and only experts in the simulation language can read the code.
be of great value in communicating the assumptions of the model to the managers (Buzacott &
Analytical models describe the system using mathematical or symbolic relationships. If the
equation methods
elton, Sadowski & Sturrock 2007:5). Buzacott &
These are then used to derive a formula or to define an
algorithm or computational procedure by which the performance measures of the system can be
s can also be used to demonstrate properties of various operating
Simulation models can be developed in several different ways. One is to try and include all the
tem from the beginning. This tends to result in long development and
debugging time, and usually makes the resultant code unreadable to all but the developer.
Another approach is to develop models of the system components and after each component
been debugged, tested and validated, link the components two at a time, test and
validate, then three at a time, test and validate and so on. Sometimes there are already existing
oo detailed for the
overall system model but often it is possible to devise ways of summarising the performance of
system component model. This
in developing models of very large
Simulation models are usually written in a simulation language such as GPSS, SLAM, SIMAN,
SIMULA or SMALLTALK and only experts in the simulation language can read the code.
Analytical models typically determine
whereas simulation models determine a time
statistical test is required to determine whether the sample is consistent with the analytical
results. Analytical models either give a formula or define a computational procedure. If they give
a computational procedure then it is necessary to consider the computational complexity of the
procedure. It is by no means uncommon for the computational requirements to increase
exponentially. This often limits the size of the system and the range of parameter values for
which the procedure can be used (Buzacott & Shanthikumar 1993:12).
2.4 Why Model?
Models can be developed for a variety of reasons, in particular:
� Understanding
“The model is used to explain why and how. The model is intended to help develop
insight. It also indicates the direction of influence of some variable on performance
by Buzacott & Shanthikumar
� Learning
“As well as providing insight a model may
about the factors that determine performance
(1993:12).
� Improvement
“The model is used to improve system design and operation. Changes in parameters
and rules can be explored, an
identified” said by Buzacott & Shanthikumar
14
Analytical models typically determine the expected value of some performance criterion,
whereas simulation models determine a time-limited sample of performance. Thus some
statistical test is required to determine whether the sample is consistent with the analytical
either give a formula or define a computational procedure. If they give
a computational procedure then it is necessary to consider the computational complexity of the
procedure. It is by no means uncommon for the computational requirements to increase
nentially. This often limits the size of the system and the range of parameter values for
which the procedure can be used (Buzacott & Shanthikumar 1993:12).
Models can be developed for a variety of reasons, in particular:
el is used to explain why and how. The model is intended to help develop
insight. It also indicates the direction of influence of some variable on performance
Buzacott & Shanthikumar (1993:12).
As well as providing insight a model may be intended to teach managers or workers
about the factors that determine performance” is said by Buzacott & Shanthikumar
The model is used to improve system design and operation. Changes in parameters
and rules can be explored, and factors critical for achieving performance targets can be
” said by Buzacott & Shanthikumar (1993:12).
the expected value of some performance criterion,
limited sample of performance. Thus some
statistical test is required to determine whether the sample is consistent with the analytical
either give a formula or define a computational procedure. If they give
a computational procedure then it is necessary to consider the computational complexity of the
procedure. It is by no means uncommon for the computational requirements to increase
nentially. This often limits the size of the system and the range of parameter values for
el is used to explain why and how. The model is intended to help develop
insight. It also indicates the direction of influence of some variable on performance”, said
be intended to teach managers or workers
” is said by Buzacott & Shanthikumar
The model is used to improve system design and operation. Changes in parameters
d factors critical for achieving performance targets can be
Understanding
Decision
making
� Optimisation
“The model is to be used to aid decisions about either the design or operation of the
system. The model has to be able to dif
and project their impact over time
Figure 5: Areas influenced by modelling
2.5 Simulation Modelling
Simulation may be defined as a technique that imitates the operation of the real
as it evolves over a period of time. There are two types of simulation models: Static and
dynamic. A static simulation model represents a system at a particular point in time. A dy
simulation model represents a system as it evolves over time.
15
Understanding Learning
Decision
makingImprovement
Optimisation
Model
The model is to be used to aid decisions about either the design or operation of the
system. The model has to be able to differentiate the effect of different courses of action
and project their impact over time”, is said by Buzacott & Shanthikumar 1993:12
modelling
Modelling
defined as a technique that imitates the operation of the real
as it evolves over a period of time. There are two types of simulation models: Static and
dynamic. A static simulation model represents a system at a particular point in time. A dy
simulation model represents a system as it evolves over time.
The model is to be used to aid decisions about either the design or operation of the
ferentiate the effect of different courses of action
”, is said by Buzacott & Shanthikumar 1993:12.
defined as a technique that imitates the operation of the real-world system
as it evolves over a period of time. There are two types of simulation models: Static and
dynamic. A static simulation model represents a system at a particular point in time. A dynamic
Winston (2004:443) said that, “
deterministic simulation contains no random variables, whereas a stochastic simulation cont
one or more random variables”. Finally, simulations may be represented by either discrete or
continuous models. A discrete simulation is one in which the state variable change only at
discrete points in time. In a continuous simulation, the state vari
time.
Simulation moves the real world to a virtual world of models, performing experiments with the
model of the system in a risk-free environment and map the best possible solution back to the
real world see Figure 6.
Figure 6: Illustrating the mapping of simulation between the real world and models
16
Winston (2004:443) said that, “Simulations can be divided into deterministic or stochastic. A
deterministic simulation contains no random variables, whereas a stochastic simulation cont
. Finally, simulations may be represented by either discrete or
continuous models. A discrete simulation is one in which the state variable change only at
discrete points in time. In a continuous simulation, the state variables change continuously over
Simulation moves the real world to a virtual world of models, performing experiments with the
free environment and map the best possible solution back to the
: Illustrating the mapping of simulation between the real world and models
Simulations can be divided into deterministic or stochastic. A
deterministic simulation contains no random variables, whereas a stochastic simulation contains
. Finally, simulations may be represented by either discrete or
continuous models. A discrete simulation is one in which the state variable change only at
ables change continuously over
Simulation moves the real world to a virtual world of models, performing experiments with the
free environment and map the best possible solution back to the
: Illustrating the mapping of simulation between the real world and models
2.5.1 Simulation Language
Computer programming is the basis of simulation
for a complex problem is a difficult and
simplify computer programming. The most widespread and readily available simula
languages include GPPS, GASP, ACSL, SIMSCRIPT, SLAM
2004:440).
2.5.2 High-Level Simulators
High-Level simulators typically operate by intuitive graphical user interface, menus and dialogs.
Selecting from available simulation
along with a dynamic graphical animation of system components as they move around and
change (Kelton, Sadowski & Sturrock 2007:5).
2.5.3 Simulation Modelling
Currently in the markets there are numerous si
own capabilities. Each tool has a different approach to simulation and their success factors vary
in more ways than one. Because of the great number of tools available only a few, which is
most applicable to this project will be analysed further.
� Arena®
“Rockwell Arena is simulation and automation software acquired by Rockwell Automation
almost a decade ago. It uses the SIMAN processor and simulation language. It is currently
in version 12.0”.
[online]http://en.wikipedia.org/wiki/Rockwell_Arena
17
Simulation Language
programming is the basis of simulation. The study and writing of the computer code
for a complex problem is a difficult and laborious task. Special purpose simulation languages
simplify computer programming. The most widespread and readily available simula
S, GASP, ACSL, SIMSCRIPT, SLAM, SIMAN and many more
Level Simulators
Level simulators typically operate by intuitive graphical user interface, menus and dialogs.
Selecting from available simulation-modelling constructs connects them and run the model
along with a dynamic graphical animation of system components as they move around and
change (Kelton, Sadowski & Sturrock 2007:5).
Modelling Tools
Currently in the markets there are numerous simulating tools available, each of which has its
own capabilities. Each tool has a different approach to simulation and their success factors vary
in more ways than one. Because of the great number of tools available only a few, which is
this project will be analysed further.
Rockwell Arena is simulation and automation software acquired by Rockwell Automation
almost a decade ago. It uses the SIMAN processor and simulation language. It is currently
(Arena Rockwell Software
http://en.wikipedia.org/wiki/Rockwell_Arena,accessed on 14May 2009)
the computer code
task. Special purpose simulation languages
simplify computer programming. The most widespread and readily available simulation
and many more (Winston
Level simulators typically operate by intuitive graphical user interface, menus and dialogs.
constructs connects them and run the model
along with a dynamic graphical animation of system components as they move around and
mulating tools available, each of which has its
own capabilities. Each tool has a different approach to simulation and their success factors vary
in more ways than one. Because of the great number of tools available only a few, which is
Rockwell Arena is simulation and automation software acquired by Rockwell Automation
almost a decade ago. It uses the SIMAN processor and simulation language. It is currently
Software – wikipedia.
accessed on 14May 2009)
Arena® combines the ease of use found in high
simulation languages and accommodates general
Microsoft® Visual Basic® programming system and even C.
Arena maintains its modelling
templates of graphical simulation
be used at any time and access to simulation language flexibility is unlimited. The Arena®
software is completely hierarchical allowing access to a whole set of simulation
constructs and capabilities, such as the SIMAN constructs together with higher
modules from other templates.
Arena® is used most effectively
involving highly sensitive changes relat
logistics, distribution, warehousing and service systems
Applications include:
• “Analysis of Six Sigma implementation, KanBan/Pull or Push systems
• “Detail analysis of warehousing, logistics or
applications”.
• “Modelling complex customer
• “Detail analysis of complex manufacturing processes that include material
intensive operations”.
http://www.arenasimulation.com/products/professional_edition.asp, accessed
18
Arena® combines the ease of use found in high-level simulators with the flexibility of
n languages and accommodates general-purpose procedural languages like the
Microsoft® Visual Basic® programming system and even C.
modelling flexibility by allowing the user to interchange between
templates of graphical simulation modelling and analysis modules. Low-level modules can
be used at any time and access to simulation language flexibility is unlimited. The Arena®
software is completely hierarchical allowing access to a whole set of simulation
es, such as the SIMAN constructs together with higher
modules from other templates.
(Kelton, Sadowski & Sturrock 2007:10)
Arena® is used most effectively “when analysing complex, medium to large
involving highly sensitive changes related to supply chain, manufacturing, processes,
logistics, distribution, warehousing and service systems”.
Analysis of Six Sigma implementation, KanBan/Pull or Push systems
Detail analysis of warehousing, logistics or transportation, military and mining
complex customer-handling activities”.
Detail analysis of complex manufacturing processes that include material
”.
(Arena® Professional Edition. [online]
http://www.arenasimulation.com/products/professional_edition.asp, accessed
level simulators with the flexibility of
purpose procedural languages like the
flexibility by allowing the user to interchange between
level modules can
be used at any time and access to simulation language flexibility is unlimited. The Arena®
software is completely hierarchical allowing access to a whole set of simulation modelling
es, such as the SIMAN constructs together with higher-level
(Kelton, Sadowski & Sturrock 2007:10)
when analysing complex, medium to large-scale projects
ed to supply chain, manufacturing, processes,
Analysis of Six Sigma implementation, KanBan/Pull or Push systems”.
transportation, military and mining
Detail analysis of complex manufacturing processes that include material-handling
(Arena® Professional Edition. [online]
http://www.arenasimulation.com/products/professional_edition.asp, accessed on 14May
2009)
Arena® exploits two Microsoft Windows® technologies that are designed to en
integration of desktop applications. The first, ActiveX® Automation, allows applications to
control each other and themselves via a programming interface. Programming languages
like C++, Visual Basic or Java is used to create the programme that c
The second technology exploited by Arena® for application integration is Visual Basic
(VBA). VBA is the same language that works with Microsoft Office, AutoCAD and Arena®.
These two technologies work together to allow Arena® to i
that support ActiveX Automation. It is possible to write Visual Basic code directly in Arena®
that automates other programmes, such as Excel, AutoCAD or Visio.
It has been noted that creating a simulation can require more time at the beginning of a
project, but quicker installations and product optimizations can reduce overall project time.
19
Arena® exploits two Microsoft Windows® technologies that are designed to en
integration of desktop applications. The first, ActiveX® Automation, allows applications to
control each other and themselves via a programming interface. Programming languages
like C++, Visual Basic or Java is used to create the programme that controls the application.
The second technology exploited by Arena® for application integration is Visual Basic
(VBA). VBA is the same language that works with Microsoft Office, AutoCAD and Arena®.
These two technologies work together to allow Arena® to integrate with other programmes
that support ActiveX Automation. It is possible to write Visual Basic code directly in Arena®
that automates other programmes, such as Excel, AutoCAD or Visio.
(Kelton, Sadowski & Sturrock 2007:420)
ating a simulation can require more time at the beginning of a
project, but quicker installations and product optimizations can reduce overall project time.
Arena® exploits two Microsoft Windows® technologies that are designed to enhance the
integration of desktop applications. The first, ActiveX® Automation, allows applications to
control each other and themselves via a programming interface. Programming languages
ontrols the application.
The second technology exploited by Arena® for application integration is Visual Basic
(VBA). VBA is the same language that works with Microsoft Office, AutoCAD and Arena®.
ntegrate with other programmes
that support ActiveX Automation. It is possible to write Visual Basic code directly in Arena®
(Kelton, Sadowski & Sturrock 2007:420)
ating a simulation can require more time at the beginning of a
project, but quicker installations and product optimizations can reduce overall project time.
� ProModel
“ProModel is discrete event simulation software, used for evaluating, planning
manufacturing, warehousing, logistics and other operational and strategic situations.
This easy-to-use software allows you to build computer models of your situation and
experiment with scenarios to find the best solution. The software include
graphical reports, providing powerful tools for analysing and optimisation
• Develop “what if” scenarios quickly
• Direct data import and export with Microsoft® Excel®
• Capture system randomness and variability by utilising over 20 statistical
types or directly import your own data.
• Distribute models to other divisions and department with run
• Automatic creation of output reports
ProModel is an easy-to-use and easy
http://www.promodel.com/products/promodel/features.asp
� ACSL
“The Advanced Continuous Simulation Language, or ACSL (pronounced “axle’), is a computer language designed for continuous system described by time
http://en.wikipedia.org/wiki/advanced_continuous_simulation_language
20
ProModel is discrete event simulation software, used for evaluating, planning
manufacturing, warehousing, logistics and other operational and strategic situations.
use software allows you to build computer models of your situation and
experiment with scenarios to find the best solution. The software include
graphical reports, providing powerful tools for analysing and optimisation”.
Develop “what if” scenarios quickly
Direct data import and export with Microsoft® Excel®
Capture system randomness and variability by utilising over 20 statistical
types or directly import your own data.
Distribute models to other divisions and department with run-time licensing
Automatic creation of output reports
use and easy-to-implement.
(ProModel Optimisation Software. [onli
http://www.promodel.com/products/promodel/features.asp, accessed on 13May 2009)
The Advanced Continuous Simulation Language, or ACSL (pronounced “axle’), is a designed for modelling and evaluating the performance of
continuous system described by time-dependent, nonlinear differential equations
(Acsl-wikipedia. [online]
http://en.wikipedia.org/wiki/advanced_continuous_simulation_language, accessed on 11May
ProModel is discrete event simulation software, used for evaluating, planning or designing
manufacturing, warehousing, logistics and other operational and strategic situations.
use software allows you to build computer models of your situation and
experiment with scenarios to find the best solution. The software includes animation and
Capture system randomness and variability by utilising over 20 statistical distribution
time licensing
(ProModel Optimisation Software. [online]
, accessed on 13May 2009)
The Advanced Continuous Simulation Language, or ACSL (pronounced “axle’), is a and evaluating the performance of
dependent, nonlinear differential equations”.
wikipedia. [online]
, accessed on 11May
2009)
ACSLX – “ACSLX is the latest software package by ACSL. ACSLX is a
execution and analysis environment for continuous dynamic systems and processes.
Simple to learn and easy to use, ACSLXtream provid
users at all level, is versatile and powerful enough to address the most challenging
simulation problems”.
The ready-to-use code blocks incorporated in ACSLX enable quick model assembly,
while powerful analysis capabilities provide quick and accurate results. Models are
specified using either familiar block diagram notation or code
CSSL (Continuous System Simulation Language)
representations are translated into C or FORTRAN language source code for
compilation and execution, ensuring the fastest possible performance and support for
extremely large or complex
(ACSLX. [online]
The Arena® package developed by Rockwell Software will be selected, investigation shows that
it is the best software for the project and
the University of Pretoria without additional expenses.
21
ACSLX is the latest software package by ACSL. ACSLX is a
execution and analysis environment for continuous dynamic systems and processes.
imple to learn and easy to use, ACSLXtream provides an intuitive environment for
users at all level, is versatile and powerful enough to address the most challenging
use code blocks incorporated in ACSLX enable quick model assembly,
while powerful analysis capabilities provide quick and accurate results. Models are
specified using either familiar block diagram notation or code-based descriptions using
(Continuous System Simulation Language) modelling language. All model
representations are translated into C or FORTRAN language source code for
compilation and execution, ensuring the fastest possible performance and support for
extremely large or complex models.
(ACSLX. [online] http://www.acslsim.com/products, accessed on 13May 2009)
The Arena® package developed by Rockwell Software will be selected, investigation shows that
roject and that a full version of this package can be provided by
the University of Pretoria without additional expenses.
ACSLX is the latest software package by ACSL. ACSLX is a modelling,
execution and analysis environment for continuous dynamic systems and processes.
es an intuitive environment for
users at all level, is versatile and powerful enough to address the most challenging
use code blocks incorporated in ACSLX enable quick model assembly,
while powerful analysis capabilities provide quick and accurate results. Models are
based descriptions using
language. All model
representations are translated into C or FORTRAN language source code for
compilation and execution, ensuring the fastest possible performance and support for
, accessed on 13May 2009)
The Arena® package developed by Rockwell Software will be selected, investigation shows that
that a full version of this package can be provided by
3 Conceptual Design of Model
3.1 Structuring the Model
A basic approach will be followed to construct the simulation model. The major
simulation study are illustrated in
simulation study, simplifies the modelling
model are made as the system is under examination. The goal of the initial model is to create a
simulation model that represents the current operation of the distribution facility.
According to Chase, Jacobs and Aquilano (2006:703) a continuous model is based on
mathematical equations with values for all points in time.
Discrete models are a process, in which changes to the state of the system occur at isolated
points in time (Kelton, Sadowski & Sturr
at Nike is a discrete simulation, events occur at specific points in time. Goods arriving at the
distribution facility are an example of a discrete event.
In a system, an object of interest is call
attributes (Winston 2004:443). The
of the attributes (one entity may have more than one attribute) attaches to that entity is the
of business units ordered by the customer,
The simulation model is also a queuing system,
station and waiting to be sent through the system.
Aquilano (2006:693).
22
Conceptual Design of Model
Structuring the Model
A basic approach will be followed to construct the simulation model. The major
simulation study are illustrated in Figure 7. This flow chart view of the major phases in
modelling process significantly, especially when changes to the
model are made as the system is under examination. The goal of the initial model is to create a
simulation model that represents the current operation of the distribution facility.
Jacobs and Aquilano (2006:703) a continuous model is based on
mathematical equations with values for all points in time.
a process, in which changes to the state of the system occur at isolated
points in time (Kelton, Sadowski & Sturrock 2007:465). The modelling of the distribution facility
at Nike is a discrete simulation, events occur at specific points in time. Goods arriving at the
distribution facility are an example of a discrete event.
In a system, an object of interest is called an entity and any properties of an entity are called
attributes (Winston 2004:443). The entity of this simulation is an order from a customer
of the attributes (one entity may have more than one attribute) attaches to that entity is the
business units ordered by the customer, i.e. footwear, apparel or equipment.
The simulation model is also a queuing system, with many orders or entities arriving at each
station and waiting to be sent through the system. Figure 7 is obtained from Chase, Jacobs and
A basic approach will be followed to construct the simulation model. The major phases in the
. This flow chart view of the major phases in
process significantly, especially when changes to the
model are made as the system is under examination. The goal of the initial model is to create a
simulation model that represents the current operation of the distribution facility.
Jacobs and Aquilano (2006:703) a continuous model is based on
a process, in which changes to the state of the system occur at isolated
of the distribution facility
at Nike is a discrete simulation, events occur at specific points in time. Goods arriving at the
ed an entity and any properties of an entity are called
entity of this simulation is an order from a customer and one
of the attributes (one entity may have more than one attribute) attaches to that entity is the type
or entities arriving at each
is obtained from Chase, Jacobs and
Figure 7: Major Phases in Simulation Study
23
: Major Phases in Simulation Study
Chase, Jacobs and Aquilano (2006:291) states a queuing system consists essentially of three
major components:
1. The source population and the ways entities arrive at the system.
2. The service system.
3. The condition of the entity exiting the system.
3.2 Conceptual Inputs and Outputs of the
The input analysis specifies the model parameters and distributions. The logic aspect of the
model, such as the entity and resource characteristics, the path an entity follows through the
system and any other activities is called structural modelling
2007:172).
The structural modelling of a system is an arrangement of the fundamental logic that describes
the system. The mathematical input that describes the logic is called quantitative
Quantitative modelling specifies the numerical nature of the resources and entities included and
not only, the distributions and model parameters modelling (Kelton, Sadowski & Sturrock
2007:172).
The inputs to the queuing model are the arrival process, the process time of each s
number of entities and the number of lines.
The structural modelling of a simple flow chart for an entity arrival is illustrated by Winston
(2004:405) in Figure 8.
24
(2006:291) states a queuing system consists essentially of three
The source population and the ways entities arrive at the system.
The condition of the entity exiting the system.
Conceptual Inputs and Outputs of the Model
The input analysis specifies the model parameters and distributions. The logic aspect of the
model, such as the entity and resource characteristics, the path an entity follows through the
system and any other activities is called structural modelling (Kelton, Sadowski & Sturrock
of a system is an arrangement of the fundamental logic that describes
the system. The mathematical input that describes the logic is called quantitative
pecifies the numerical nature of the resources and entities included and
not only, the distributions and model parameters modelling (Kelton, Sadowski & Sturrock
The inputs to the queuing model are the arrival process, the process time of each s
number of entities and the number of lines.
of a simple flow chart for an entity arrival is illustrated by Winston
(2006:291) states a queuing system consists essentially of three
The input analysis specifies the model parameters and distributions. The logic aspect of the
model, such as the entity and resource characteristics, the path an entity follows through the
(Kelton, Sadowski & Sturrock
of a system is an arrangement of the fundamental logic that describes
the system. The mathematical input that describes the logic is called quantitative modelling.
pecifies the numerical nature of the resources and entities included and
not only, the distributions and model parameters modelling (Kelton, Sadowski & Sturrock
The inputs to the queuing model are the arrival process, the process time of each station, the
of a simple flow chart for an entity arrival is illustrated by Winston
Figure 8: Flow chart of an Arrival
The values of the inputs to the simulation are determined by using Microsoft Excel®. Excel®
spreadsheets are used as a user
which will be used in the simulation model.
By linking these spreadsheets with the model, the input values can be altered for different
scenarios without having to change the values in the simulation model itself. All the
spreadsheets are linked with each other, providing a relationship between interrelated cells.
This makes it easier for a person with no knowledge of how Arena® operates to run different
scenarios and draw data from the model.
25
values of the inputs to the simulation are determined by using Microsoft Excel®. Excel®
spreadsheets are used as a user-interface tool to assign values to attributes and variables,
which will be used in the simulation model.
By linking these spreadsheets with the model, the input values can be altered for different
scenarios without having to change the values in the simulation model itself. All the
eets are linked with each other, providing a relationship between interrelated cells.
This makes it easier for a person with no knowledge of how Arena® operates to run different
scenarios and draw data from the model.
values of the inputs to the simulation are determined by using Microsoft Excel®. Excel®
ttributes and variables,
By linking these spreadsheets with the model, the input values can be altered for different
scenarios without having to change the values in the simulation model itself. All the
eets are linked with each other, providing a relationship between interrelated cells.
This makes it easier for a person with no knowledge of how Arena® operates to run different
The conceptual outputs of the queuing system are typified by a specific probability distribution,
which governs a customer’s service time. The random variable input data generates random
variability in the output data. The most important output statistic of a simulation model will
include the following performance measures (Mehrotra & Fama
• Abandonment Statistics
• Queue Statistics
• Volume Statistics
The system will be measured by several KPI’s (Key Performance Indicators)
used to measure the performance of
important KPI’s will include the following
1. Resource Utilisation (Picker Utilisation and Packer Utilisation)
time that a resource is truly being used in relation to the time that it is available.
2. Throughput Time – It includes the time that the unit is being worked on together with
the time that the unit spends waiting in a queue
through the system).
3. Time in Process – The time a unit takes to be completed in a specific process e.g. the
time it takes a unit to pass through the picking phase.
26
queuing system are typified by a specific probability distribution,
which governs a customer’s service time. The random variable input data generates random
variability in the output data. The most important output statistic of a simulation model will
ude the following performance measures (Mehrotra & Fama 2003:137).
he system will be measured by several KPI’s (Key Performance Indicators). These KPI’s are
measure the performance of the different scenarios against each other.
will include the following as stated by Chase, Jacobs & Aquilano
(Picker Utilisation and Packer Utilisation) – It is the ratio of the
e is truly being used in relation to the time that it is available.
It includes the time that the unit is being worked on together with
the time that the unit spends waiting in a queue (The average time a unit takes to move
The time a unit takes to be completed in a specific process e.g. the
time it takes a unit to pass through the picking phase.
queuing system are typified by a specific probability distribution,
which governs a customer’s service time. The random variable input data generates random
variability in the output data. The most important output statistic of a simulation model will
. These KPI’s are
the different scenarios against each other. The more
Chase, Jacobs & Aquilano (2006:164).
It is the ratio of the
e is truly being used in relation to the time that it is available.
It includes the time that the unit is being worked on together with
(The average time a unit takes to move
The time a unit takes to be completed in a specific process e.g. the
4 Data Collection and Analysis
One of the very early steps in planning a simulation project should be to identify what data is
needed to support the model. The availability and quality of data can influence the
approach taken and the level of detail captured in the model
2007:174).
Accourding to Kelton, Sadowski & Sturrock (2007:175) a benefit of developing an early
understanding of the quality of your input data is that it can help you decide how much detail to
incorporate in the model logic. The
study are only as reliable as the model and its inputs.
Input Data Analyzer is included in the Arena® software, it automatically fits existing data values
to the best statistical probability
data and provides an estimate of the parameter values and an expression of the model (Kelton,
Sadowski & Sturrock 2007:176).
4.1 Input Data and Analysis
Data will be received by intensive time
retrieved from the Microsoft Access® data base. It is then transferred to Microsoft Ex
is more accessible and the data is analysed.
4.1.1 Order Information
Each order generated by a customer has cert
ordered, the quantity of each item and the delivery date. Each order is then classified into three
different channels, the IDP (Integrated Delivery Plan), Non IDP and the Priorities. The channels
differ in the delivery dates, how urgent a certain order is. The IDP has a fixed despatch date and
planning is done around that date. The Non IDP is the most flexible channel it has a shipping
27
Data Collection and Analysis
One of the very early steps in planning a simulation project should be to identify what data is
needed to support the model. The availability and quality of data can influence the
approach taken and the level of detail captured in the model (Kelton, Sadowski & Sturrock
Kelton, Sadowski & Sturrock (2007:175) a benefit of developing an early
understanding of the quality of your input data is that it can help you decide how much detail to
incorporate in the model logic. The results and recommendations present, from the simulation
study are only as reliable as the model and its inputs.
Input Data Analyzer is included in the Arena® software, it automatically fits existing data values
to the best statistical probability distributions. The Input Data Analyzer fits a distribution to the
data and provides an estimate of the parameter values and an expression of the model (Kelton,
Sadowski & Sturrock 2007:176).
and Analysis
Data will be received by intensive time studies on the relevant stations. The input data is
retrieved from the Microsoft Access® data base. It is then transferred to Microsoft Ex
is more accessible and the data is analysed.
by a customer has certain data attached to it. The types
ordered, the quantity of each item and the delivery date. Each order is then classified into three
different channels, the IDP (Integrated Delivery Plan), Non IDP and the Priorities. The channels
n the delivery dates, how urgent a certain order is. The IDP has a fixed despatch date and
planning is done around that date. The Non IDP is the most flexible channel it has a shipping
One of the very early steps in planning a simulation project should be to identify what data is
needed to support the model. The availability and quality of data can influence the modelling
on, Sadowski & Sturrock
Kelton, Sadowski & Sturrock (2007:175) a benefit of developing an early
understanding of the quality of your input data is that it can help you decide how much detail to
from the simulation
Input Data Analyzer is included in the Arena® software, it automatically fits existing data values
distributions. The Input Data Analyzer fits a distribution to the
data and provides an estimate of the parameter values and an expression of the model (Kelton,
studies on the relevant stations. The input data is
retrieved from the Microsoft Access® data base. It is then transferred to Microsoft Excel®, which
n data attached to it. The types of goods that is
ordered, the quantity of each item and the delivery date. Each order is then classified into three
different channels, the IDP (Integrated Delivery Plan), Non IDP and the Priorities. The channels
n the delivery dates, how urgent a certain order is. The IDP has a fixed despatch date and
planning is done around that date. The Non IDP is the most flexible channel it has a shipping
window of delivery date plus 30 days. T
in today it must be despatched the next day, making it the least flexible channel.
the most important channel for the world cup season.
4.1.2 Transport Time of a Picker per Order
The Transport time of a picker per ord
from the one bin location to the next in order to pick the units
consuming as the order could contain any of the 30,000+ products Nike offer.
Each unit has a specified location that is entered into the system. The levels of the pick face is
divided into blocks, level A is divided into 3x9 blocks and levels B,C and D is divided into 3x12
blocks. The blocks are defined by an
and the distances from one block to the next. The locations of the units are then incorporated
into one of these blocks, thus each un
are linked via there reference to a blo
The time it takes a picker to walk from one block to the next is calculated through time studies
(see Figure 9). Level A differs from levels B
locations on these levels.
The time to walk from one block to the next or to any other block is calculated in Microsoft
Access®. A formula is used where the times are multiplied if the picker either walks in
direction or in the y-direction. The formula differs from level A to the others
formula is multiplied with. E.g. level A is multiplied in the x
multiplied with 14 (see Figure 9 for the values).
28
livery date plus 30 days. The Priorities is the most urgent orders, if an order comes
ust be despatched the next day, making it the least flexible channel.
the most important channel for the world cup season.
Transport Time of a Picker per Order
The Transport time of a picker per order is the time a picker takes to walk on a specific level
from the one bin location to the next in order to pick the units to fulfil the order
consuming as the order could contain any of the 30,000+ products Nike offer.
specified location that is entered into the system. The levels of the pick face is
, level A is divided into 3x9 blocks and levels B,C and D is divided into 3x12
ks. The blocks are defined by an X-Y co-ordination in order to calculate the size of a block
and the distances from one block to the next. The locations of the units are then incorporated
into one of these blocks, thus each unit belongs to a specific block on a specific level.
to a block.
The time it takes a picker to walk from one block to the next is calculated through time studies
). Level A differs from levels B, C and D because of the greater number of
The time to walk from one block to the next or to any other block is calculated in Microsoft
Access®. A formula is used where the times are multiplied if the picker either walks in
direction. The formula differs from level A to the others by;
formula is multiplied with. E.g. level A is multiplied in the x-direction with 11 and levels B, C, D is
for the values).
nt orders, if an order comes
ust be despatched the next day, making it the least flexible channel. This channel is
er is the time a picker takes to walk on a specific level
order. This is time
specified location that is entered into the system. The levels of the pick face is
, level A is divided into 3x9 blocks and levels B,C and D is divided into 3x12
e the size of a block
and the distances from one block to the next. The locations of the units are then incorporated
it belongs to a specific block on a specific level. The units
The time it takes a picker to walk from one block to the next is calculated through time studies
and D because of the greater number of storage
The time to walk from one block to the next or to any other block is calculated in Microsoft
Access®. A formula is used where the times are multiplied if the picker either walks in the x-
by; the number the
direction with 11 and levels B, C, D is
Figure 9: Walking time from one block to next
The travel times are then exported to Microsoft Excel® where the data are pivoted into a table
for travel times of level A and a separate table for levels B,
15 and Table 16). If picking at a certain level does occur, the transport time is calculated in
Microsoft Excel® using the Index and Match formulas. The formula is an “if” statement, it looks
on what level the picking must occur and uses the level A table or the level B, C, D
“if” statement the Index function is used marking the table that needs to be used. In combination
with Index function the Match function is used in order the match
starting and ending locations), the picker must
one block to the next block where the picker must pick next.
After the travel times between the blocks, in order to pick units,
time to the conveyor is calculated
29
: Walking time from one block to next
then exported to Microsoft Excel® where the data are pivoted into a table
for travel times of level A and a separate table for levels B, C and D (see Appendix
picking at a certain level does occur, the transport time is calculated in
Microsoft Excel® using the Index and Match formulas. The formula is an “if” statement, it looks
on what level the picking must occur and uses the level A table or the level B, C, D
“if” statement the Index function is used marking the table that needs to be used. In combination
with Index function the Match function is used in order the match, from where and
the picker must walk. This value is the transport time from the
one block to the next block where the picker must pick next.
between the blocks, in order to pick units, are calculated
calculated (after all picks is completed and the units must be taken to the
then exported to Microsoft Excel® where the data are pivoted into a table
ppendix 10.1, Table
picking at a certain level does occur, the transport time is calculated in
Microsoft Excel® using the Index and Match formulas. The formula is an “if” statement, it looks
on what level the picking must occur and uses the level A table or the level B, C, D table. In the
“if” statement the Index function is used marking the table that needs to be used. In combination
where and to where (the
This value is the transport time from the
are calculated then the travel
ks is completed and the units must be taken to the
conveyor). There are three blocks in the x
irrelative in what y-block the picker is, is the block number in the x
E.g. If the picker stands in the second block in the x
conveyor is 2*7.5 = 15sec. This is calculated for every order on a specific floor.
The total transport time for a picker is the sum of the transport times. These
expanded into their different levels (A, B, C, D) by placing them into a matrix form (see
Appendix 10.1, Table 17).
4.1.3 Quantity of a Business Unit Picked
An order contains the quantity of certain
Access® database then finds the loc
quantity units, type of business unit (footwear or other) and the level it is situated in. A matrix
table is then compiled of the information in Microsoft Excel® from the data in the Microsoft
Access® database. The matrix table
quantity that must be picked on a level (s
computation between Microsoft Access® and Excel® is done automatically
4.1.4 Number to be batched
An order can have units that need
need to be picked on different levels,
different levels. This split in the order must be batched toge
boxes. The number to be batched calculates
be batched must be equal to the number of splits
not be despatched until the number of splits are all together
to assure that all orders are complete when
30
. There are three blocks in the x-direction, the time it takes to walk to the conveyor
block the picker is, is the block number in the x-direction multiplied with 7.5.
If the picker stands in the second block in the x-direction then the transport time to the
conveyor is 2*7.5 = 15sec. This is calculated for every order on a specific floor.
The total transport time for a picker is the sum of the transport times. These
expanded into their different levels (A, B, C, D) by placing them into a matrix form (see
Quantity of a Business Unit Picked
An order contains the quantity of certain products and the type of business units. The Microsoft
Access® database then finds the locations of the products that are ordered. It captures the
quantity units, type of business unit (footwear or other) and the level it is situated in. A matrix
of the information in Microsoft Excel® from the data in the Microsoft
The matrix table states the order number; the type of business units and the
that must be picked on a level (see Appendix 10.1, Table 18). Each order differs and all
computation between Microsoft Access® and Excel® is done automatically for each order
hed
units that need to be picked on different levels. If an order has
need to be picked on different levels, it is then split by the units that need
different levels. This split in the order must be batched together when the order is packed into
number to be batched calculates how many splits an order has and the number to
be batched must be equal to the number of splits, this is illustrated in Figure 10
not be despatched until the number of splits are all together (see Appendix 10.1
orders are complete when despatched.
to walk to the conveyor
direction multiplied with 7.5.
direction then the transport time to the
The total transport time for a picker is the sum of the transport times. These times are then
expanded into their different levels (A, B, C, D) by placing them into a matrix form (see
and the type of business units. The Microsoft
that are ordered. It captures the
quantity units, type of business unit (footwear or other) and the level it is situated in. A matrix
of the information in Microsoft Excel® from the data in the Microsoft
business units and the
). Each order differs and all
for each order.
an order has units that
to be picked at
ther when the order is packed into
and the number to
10. The order will
10.1, Table 19). It is
Figure 10: Illustration of the split of an Order
4.1.5 Distance Conveyor travel in Picking
On the side of every level is a conveyor (see
order, the tote (picked units) is then placed
packing. The conveyor moves from level to level
Each order ends on a different place on a level. The picker then walks straight
picked locations to the conveyor. The
has an influence on the distance
on the time it takes a tote to travel to packing.
If a tote is placed at position A,
point A to the end of level A, shown as a blue arrow
then the tote must move the distance on level D and the entire distance on level C
green arrow. Thus the tote at point B will travel longer to get to packing than point A.
31
split of an Order
Distance Conveyor travel in Picking
On the side of every level is a conveyor (see Figure 9), when a picker is finished picking
is then placed on the conveyor. The conveyor conveys the totes to
The conveyor moves from level to level as illustrated in Figure 11 by the red arrows
Each order ends on a different place on a level. The picker then walks straight
to the conveyor. The location of where the totes are placed on the conveyor
has an influence on the distance that the tote must travel to packing; this has a direct influence
on the time it takes a tote to travel to packing.
in Figure 11, then the tote only has to travel the distance from
, shown as a blue arrow. If the tote is placed at point B in
then the tote must move the distance on level D and the entire distance on level C
at point B will travel longer to get to packing than point A.
a picker is finished picking an
on the conveyor. The conveyor conveys the totes to
by the red arrows.
Each order ends on a different place on a level. The picker then walks straight from the last
placed on the conveyor
this has a direct influence
, then the tote only has to travel the distance from
ed at point B in Figure 11,
then the tote must move the distance on level D and the entire distance on level C, shown as a
at point B will travel longer to get to packing than point A.
Figure 11: Illustration of Conveyor Movement
The location of the last product on the pick list is used to determine the travelling
level A the y co-ordination is 9 blocks and on levels B, C, D there are
Figure 12, the distance from the la
shown as a green arrow, and the distance of which the conveyor is
red arrow, is added together (see
used.
32
: Illustration of Conveyor Movement on Picking Levels
The location of the last product on the pick list is used to determine the travelling
ordination is 9 blocks and on levels B, C, D there are 12 blocks as illustrated in
he distance from the last pick point to the end of the level is calculated
and the distance of which the conveyor is longer, which
(see Figure 12).In essence the y co-ordinate of the final location is
The location of the last product on the pick list is used to determine the travelling distance. If on
12 blocks as illustrated in
st pick point to the end of the level is calculated, which is
longer, which is shown as a
ordinate of the final location is
Figure 12: Level Layout
This calculation is done in Microsoft Excel® wi
picker is situated on. If the picker is on level A then
level A. If the picker is on levels B,
those levels contain 12 blocks. E.g.
the distance marked by the letter A
conveyor is longer, marked with the red arrow
last pick point to the end of the conveyor.
The distances used in the “if” statements are shown in the
33
This calculation is done in Microsoft Excel® with 21 “if” statements. It determines
. If the picker is on level A then a certain distance is used for the 9 blocks on
level A. If the picker is on levels B, C, D then a different distance, than level A is used, because
12 blocks. E.g. If the picker is on level C (see Figure 12) in block 3
the distance marked by the letter A is calculate and then added with the distance which th
, marked with the red arrow. The total y co-ordination is calculated from the
last pick point to the end of the conveyor.
The distances used in the “if” statements are shown in the Table 1.
determines the level the
used for the 9 blocks on
n level A is used, because
) in block 3-5 then
is calculate and then added with the distance which the
ordination is calculated from the
Table 1: Distance Conveyor travel from each block
Block Number in
y co-ordination
1
2
3
4
5
6
7
8
9
10
11
12
After conducting a few measurements on the conveyors system it was concluded that all
conveyor travel at a constant speed.
obtain the time the conveyor takes to move from a level in picking to packing.
the different level is shown in a matrix
Appendix 10.1, Table 20.
4.1.6 Process Time Picking
The process time is the time a picker takes to pick a number of units of a specific business unit
(footwear, apparel or equipment). Intensive time studies were done and
each station. The time to pick footwear differs to the other business units
34
: Distance Conveyor travel from each block
Distance conveyor travel (m)
Block Number in
Level A Level B Level C
6 70 5
12 66 9
18 61 14
24 56 18
30 52 23
36 47 28
42 43 32
48 38 37
54 33 41
29 46
24 51
20 55
After conducting a few measurements on the conveyors system it was concluded that all
conveyor travel at a constant speed. The distance is then multiplied with the speed
obtain the time the conveyor takes to move from a level in picking to packing.
the different level is shown in a matrix form and these distances are read into the
Process Time Picking
The process time is the time a picker takes to pick a number of units of a specific business unit
(footwear, apparel or equipment). Intensive time studies were done and data was obtained from
each station. The time to pick footwear differs to the other business units due to
Level C
After conducting a few measurements on the conveyors system it was concluded that all
speed in order to
obtain the time the conveyor takes to move from a level in picking to packing. The distance of
read into the simulation see
The process time is the time a picker takes to pick a number of units of a specific business unit
data was obtained from
due to the irregular
size difference in apparel and equipment
time to pick thus they are combined into
from the time studies were analysed
Table 2: Processing Times for Picking
Process Time (in sec)
Other
Footwear
Using these results mathematical expression
(Appendix 10.2.1 and 10.2.2) which indicates
and maximum as stated in Table
change in the simulation model, thus they are user defined.
4.1.7 Process Time Packing
The process packing time includes all the different business units, no distinction can be made
between footwear and other. The data obtained for packing follows a
developed in Arena® Input Analyzer (
average and the standard deviation
will automatically change in the simulation model, thus they are user defined.
Table 3: Processing Times for Packing
Packing Time
35
in apparel and equipment. Apparel and equipment takes the same amount of
time to pick thus they are combined into a business unit called “Other”. The results obtained
were analysed.
: Processing Times for Picking in sec/unit
Process Time (in sec) Minimum Average Maximum
1 19
3 11
results mathematical expressions were developed in Arena® Input Analyzer
which indicates triangular distributions with the minimum, average
Table 2. These values may be changed and will automatically
change in the simulation model, thus they are user defined.
Process Time Packing
The process packing time includes all the different business units, no distinction can be made
between footwear and other. The data obtained for packing follows a mathematical expression
developed in Arena® Input Analyzer (Appendix 10.2.3) which is a normal distribution
average and the standard deviation are indicated in Table 3. These values may be changed and
will automatically change in the simulation model, thus they are user defined.
: Processing Times for Packing
Average (in sec) Std Dev
287 98.2
. Apparel and equipment takes the same amount of
The results obtained
Maximum
48
26
developed in Arena® Input Analyzer
with the minimum, average
These values may be changed and will automatically
The process packing time includes all the different business units, no distinction can be made
mathematical expression
normal distribution. The
These values may be changed and
4.2 Soft Data
The soft data include all the data relating to the workforce and personnel. The company policies
govern the staffing regulations, such as the number of staff members with their respective skill
levels, daily shift hours and weekly schedules.
Figure 13 is adapted in combining information from Buzacott & Shanthikumar (1993), Kelton,
Sadowski & Sturrock (2007) and Chase, Jacobs and Aquilano (2006) to illu
involved in the simulation study.
Figure 13: Data input to Simulation Study
36
The soft data include all the data relating to the workforce and personnel. The company policies
govern the staffing regulations, such as the number of staff members with their respective skill
levels, daily shift hours and weekly schedules.
is adapted in combining information from Buzacott & Shanthikumar (1993), Kelton,
Sadowski & Sturrock (2007) and Chase, Jacobs and Aquilano (2006) to illu
: Data input to Simulation Study
The soft data include all the data relating to the workforce and personnel. The company policies
govern the staffing regulations, such as the number of staff members with their respective skill
is adapted in combining information from Buzacott & Shanthikumar (1993), Kelton,
Sadowski & Sturrock (2007) and Chase, Jacobs and Aquilano (2006) to illustrate the data
5 Depiction of Simulation Model
The development of the simulation model is seen as the most difficult and critical part because
of the detailed description and conceptual translation of the system. Logic relations combines
with mathematical relations must represent the important attribut
simulation model is divided into four
5.1 Conceptual Translation
The conceptual translation of a model lays an essential foundation in the progress or design
phase and is also referred to as identification or the ‘piec
translating a model is to, formulate the model design and approach before one actually
the model in Arena®. The justification behind this is to ensure that all mechanisms and ‘building
components’ are considered. Some of
structure or constraints, the type of analysis to be performed, the type of animation required and
replication length, just to name a few.
5.1.1 Entities
Entities are the dynamic objects of the simulatio
are affected by other entities and the state of the system and affect the output performance
measures (Kelton, Sadowski & Sturrock 2007:20).
A single entity is created at time zero merely to assign a set of at
model. A Read/Write module is used to obtain all
Excel® spreadsheets.
The second entity which is created receives the attributes assigned to it.
replicated and each replication is
5.1.2 Attributes
An attribute is a common characteristic of all entities, but with a specific value that can differ
from one entity to another (Kelton, Sadowski & Sturrock 2007:2
37
Depiction of Simulation Model
The development of the simulation model is seen as the most difficult and critical part because
of the detailed description and conceptual translation of the system. Logic relations combines
with mathematical relations must represent the important attributes of the processes.
tion model is divided into four activity areas.
Conceptual Translation
The conceptual translation of a model lays an essential foundation in the progress or design
phase and is also referred to as identification or the ‘pieces of the model’. Conceptual
formulate the model design and approach before one actually
the model in Arena®. The justification behind this is to ensure that all mechanisms and ‘building
components’ are considered. Some of the things that one should take into account are the data
structure or constraints, the type of analysis to be performed, the type of animation required and
replication length, just to name a few.
Entities are the dynamic objects of the simulation. They move around, change status, affect and
e affected by other entities and the state of the system and affect the output performance
measures (Kelton, Sadowski & Sturrock 2007:20).
A single entity is created at time zero merely to assign a set of attributes and variables to the
model. A Read/Write module is used to obtain all the attributes and variables from
The second entity which is created receives the attributes assigned to it.
replication is specified as an order made by a customers.
An attribute is a common characteristic of all entities, but with a specific value that can differ
from one entity to another (Kelton, Sadowski & Sturrock 2007:21).
The development of the simulation model is seen as the most difficult and critical part because
of the detailed description and conceptual translation of the system. Logic relations combines
es of the processes. The
The conceptual translation of a model lays an essential foundation in the progress or design
es of the model’. Conceptual
formulate the model design and approach before one actually builds
the model in Arena®. The justification behind this is to ensure that all mechanisms and ‘building
the things that one should take into account are the data
structure or constraints, the type of analysis to be performed, the type of animation required and
n. They move around, change status, affect and
e affected by other entities and the state of the system and affect the output performance
tributes and variables to the
the attributes and variables from the Microsoft
The second entity which is created receives the attributes assigned to it. This entity gets
An attribute is a common characteristic of all entities, but with a specific value that can differ
By attaching multiple attribute values to each en
of each activity will be listed independently, since their values are determined by an Excel®
spreadsheet. All the attributes needed
Table 4.
Table 4: List of Attributes
Attribute Name
TransTime_A
TransTime_B
TransTime_C
TransTime_D
the picker to walk from one pick
station to the next on a specific
Order#
#beBatched must be batched, if the order is
Qty_LA_Other
Qty_LA_FW
Qty_LB_Other
Qty_LB_FW
Qty_LC_Other
Qty_LC_FW
Qty_LD_Other
Qty_LD_FW
ConvDist_LA
conveyor to the end of the level
ConvDist_LB
ConvDist_LC
ConvDist_LD
38
attribute values to each entity, each entity becomes unique.
activity will be listed independently, since their values are determined by an Excel®
needed to perform the outbound process flow model are listed
Description Purpose
Time (Transport time) it takes
the picker to walk from one pick
station to the next on a specific
level.
To measure the time it takes
to walk, a specific distance by
that operator for an order.
Each order has a specific
number which is call the order
number.
In order to rectify what
customer placed what order.
It indicated how many totes
must be batched, if the order is
split in the picking phase.
To make sure that all the totes
that have been split is
together for the
The number of Other/FW units
that needs to be picked at a
specific level (Level A, B, C or
D)
It is used to be multiplied by
the time to pick that specific
unit. In order to get the total
processes picking time.
The distance the conveyor
needs to travel from where the
picker places the units on the
conveyor to the end of the level
see Figure 12.
To calculate the total time the
conveyor takes to transport
the units to the next station.
The distance is multiplied by
the speed of the conv
each entity becomes unique. The attributes
activity will be listed independently, since their values are determined by an Excel®
model are listed in
Purpose
To measure the time it takes
to walk, a specific distance by
that operator for an order.
In order to rectify what
customer placed what order.
To make sure that all the totes
that have been split is
together for the same order.
It is used to be multiplied by
the time to pick that specific
unit. In order to get the total
processes picking time.
To calculate the total time the
yor takes to transport
to the next station.
ce is multiplied by
of the conveyor.
5.1.3 Variables
A variable is a piece of information that reflects some characteristic of your system, regardless
of how or what kind of entities might be around.
model (Kelton, Sadowski & Sturrock 2007:21).
Table 5: List of Variables
Variable Name
PT_Other_Min
PT_Other_Avg
PT_Other_Max
PT_FW_Min
PT_FW_Avg
PT_FW_Max
PT_Pack_Avg
PT_Pack_StdDev
Input_Taller
Input_Taller + 1
Max#Order
39
A variable is a piece of information that reflects some characteristic of your system, regardless
of how or what kind of entities might be around. There can be many different variables in a
model (Kelton, Sadowski & Sturrock 2007:21).
Description Purpose
The range of the process
time (picking time). The
process time, is the time a
picker uses to pick a single
unit of type Other.
It is used to calculate the time
a picker takes to pick a
specified number of unit and
of type Other. The quantity is
multiplied by process time.
The range of the process
time (picking time). The
process time, is the time a
picker uses to pick a single
unit of type FW (Footwear).
It is used to calculate the time
a picker takes to pick a
specified number of unit and
of type FW. The quantity is
multiplied by process time.
It is the average packing
process time and its standard
deviation.
The packing follows a normal
distribution. The average and
the standard deviation are
used as inputs for the
distribution.
It is a variable that is given
when an entity pass.
It is used to form the logic
that a specified number of
orders are released from the
pick pool for picking.
It is the total number of
orders placed by the pick
pool (the number of the last
order).
It is used to find the last entity
(order) in the simulation
model.
A variable is a piece of information that reflects some characteristic of your system, regardless
There can be many different variables in a
Purpose
It is used to calculate the time
a picker takes to pick a
specified number of unit and
of type Other. The quantity is
multiplied by process time.
It is used to calculate the time
a picker takes to pick a
specified number of unit and
of type FW. The quantity is
multiplied by process time.
The packing follows a normal
distribution. The average and
the standard deviation are
used as inputs for the
distribution.
It is used to form the logic
that a specified number of
orders are released from the
pick pool for picking.
is used to find the last entity
(order) in the simulation
model.
The variables used are user-defined and the values are determined by Excel® spreadsheets.
The variables used to perform all the actions in the model are listed in
5.1.4 Resources
Resources represent things like personnel or equipment. An entity seizes (
a resource when available and release
2007:22).
The following resources were identified in the model operations:
Table 6: List of Resources
Resource Name
Picker
Packer A-J
The current system contains 8 picker resource units for the picking and
station, there are 10 pack stations.
40
defined and the values are determined by Excel® spreadsheets.
The variables used to perform all the actions in the model are listed in Table 5.
Resources represent things like personnel or equipment. An entity seizes (a number of
a resource when available and releases it (or them) when finished (Kelton, Sadowski & Sturrock
The following resources were identified in the model operations:
Resource Name Description
It is a person who receives an order
that needs to be picked by hand. The
picker walks to every station and pick
the specific amount of units and place it
in a tote, when finished the tote is
placed on the conveyor.
J
It is a person who packs all the single
units from the totes into boxes for
shipment.
picker resource units for the picking and 2 packers
pack stations.
defined and the values are determined by Excel® spreadsheets.
a number of units of)
s it (or them) when finished (Kelton, Sadowski & Sturrock
It is a person who receives an order
that needs to be picked by hand. The
picker walks to every station and pick
the specific amount of units and place it
in a tote, when finished the tote is
It is a person who packs all the single
boxes for
packers for every pack
5.1.5 Queues
The purpose of a queue is when an entity can not move on because it needs to seize a unit
resource that is tied up by another entity; it needs to be place to wait. (Kelton, Sadowski &
Sturrock 2007:22).That entity will wait in the queue until a resources is available.
The queues defined in the simulation model are listed
Table 7: List of Queues
Queue Name
Transport between Station_LA.Queue
Transport between Station_LB.Queue
Transport between Station_LC.Queue
Transport between Station_LD.Queue
Batch Orders same OrderNumb.Queue
Packer Station_A.Queue
Packer Station_B.Queue
Packer Station_C.Queue
Packer Station_D.Queue
Packer Station_E.Queue
Packer Station_F.Queue
Packer Station_G.Queue
Packer Station_H.Queue
Packer Station_I.Queue
Packer Station_J.Queue
41
The purpose of a queue is when an entity can not move on because it needs to seize a unit
resource that is tied up by another entity; it needs to be place to wait. (Kelton, Sadowski &
Sturrock 2007:22).That entity will wait in the queue until a resources is available.
imulation model are listed in Table 7.
Description
Transport between Station_LA.Queue The waiting of a picker resource at a specific
level (A, B, C, D). In order to seize it for
picking at that level.
Transport between Station_LB.Queue
Transport between Station_LC.Queue
Transport between Station_LD.Queue
Batch Orders same OrderNumb.Queue
If an order is split at picking it needs
batched together to ensure that the full order
(not just a part of the order) is despatched to
the customer. It is the waiting of an entity (part
of an order) for the rest of the order to be
completed.
Packer Station_A.Queue
It is the waiting of an entity at the pack station,
for a specific resource, in order to pack the
complete order.
Packer Station_B.Queue
Packer Station_C.Queue
Packer Station_D.Queue
Packer Station_E.Queue
Packer Station_F.Queue
Packer Station_G.Queue
Station_H.Queue
Packer Station_I.Queue
Packer Station_J.Queue
The purpose of a queue is when an entity can not move on because it needs to seize a unit of a
resource that is tied up by another entity; it needs to be place to wait. (Kelton, Sadowski &
Sturrock 2007:22).That entity will wait in the queue until a resources is available.
Description
The waiting of a picker resource at a specific
level (A, B, C, D). In order to seize it for
picking at that level.
If an order is split at picking it needs to be
batched together to ensure that the full order
(not just a part of the order) is despatched to
the customer. It is the waiting of an entity (part
of an order) for the rest of the order to be
entity at the pack station,
for a specific resource, in order to pack the
complete order.
5.1.6 Conveyors
Long distance transportation of product is facilitated through the use of belt conveyors systems.
The following conveyors are activities on their own. They were identified to convey totes from
the picking station to the packing station and then from the packing station to the despatching
area. The conveyors in the picking and packing stations move at a
Table 8: Conveyors
Conveyor Name
Conveyor Picking
Conveyor Packing
5.1.7 Picker Travel Pattern
Each order that a picker receive
unique to that specific order) this is shown in
pattern is calculated so that the picking of the units occur in the least amount of walking
distance. The travelling time is proportional to the
the next.
To obtain the travel time of the picker from each station, the distance must be calculated
stated in section 4.1.2. The calculation is done in
model.
Level A is divided into 27 blocks
from the centre of the one block to the next is known and this includes all the blocks on all the
levels. This gives the distance travel by the picker from the one block to the next.
42
Long distance transportation of product is facilitated through the use of belt conveyors systems.
The following conveyors are activities on their own. They were identified to convey totes from
the picking station to the packing station and then from the packing station to the despatching
. The conveyors in the picking and packing stations move at a constant speed
Conveyor Name Description
Conveyor Picking
It is the time a conveyor uses to
transport totes from the picking
stations to the packing stations.
Conveyor Packing
It is the time a conveyor uses to
convey the boxes from the packing
stations to the despatching station.
a picker receive has a unique travel pattern on that level (the travel pattern is
this is shown in Figure 14 and described in section 4.1.2
so that the picking of the units occur in the least amount of walking
is proportional to the distance travelled from one pick
To obtain the travel time of the picker from each station, the distance must be calculated
. The calculation is done in Microsoft Excel® and read in
blocks and levels B, C and D is divided into 36 blocks.
from the centre of the one block to the next is known and this includes all the blocks on all the
distance travel by the picker from the one block to the next.
Long distance transportation of product is facilitated through the use of belt conveyors systems.
The following conveyors are activities on their own. They were identified to convey totes from
the picking station to the packing station and then from the packing station to the despatching
speed.
It is the time a conveyor uses to
transport totes from the picking
stations to the packing stations.
It is the time a conveyor uses to
convey the boxes from the packing
stations to the despatching station.
(the travel pattern is
and described in section 4.1.2. This
so that the picking of the units occur in the least amount of walking
from one pick location to
To obtain the travel time of the picker from each station, the distance must be calculated as
in to the simulation
and levels B, C and D is divided into 36 blocks. The distance
from the centre of the one block to the next is known and this includes all the blocks on all the
distance travel by the picker from the one block to the next. The system
knows what type of units is stored in each block and a
sequential distance for each order. This gives the travelling distance of a picker betwee
stations and that travelling distance is used to
Figure 14: Picker Travel Pattern
43
stored in each block and a walking pattern is developed for the
order. This gives the travelling distance of a picker betwee
stations and that travelling distance is used to calculate the travelling time of the picker.
walking pattern is developed for the
order. This gives the travelling distance of a picker between
the travelling time of the picker.
6 Simulation Model Construction
The simulation model construction can be very time consuming, depending on its detail and
complexity. The best approach is to start with a simple model and increasing the detail and
complexity as system knowledge
methods used are discussed in this section.
6.1 Model Activity Areas
The simulation model construction is divided into several activity areas to better understand and
illustrate the logic of the model. The activity areas are as follows:
1. Order information generation
2. Pick face
3. Pack stations
4. Despatch and Output transfer
6.1.1 Order Information Generation
In this activity area two entities
variables and the second entity is
The first entity is created at time zero and runs through the read modules
the different variables as in the picking processing time of the differe
processing time and the total number of order
that are imported from the Microsoft Excel® spreadsheets. The variables are then incorporated
into the different distributions for the picking and packing processing times, these distributions
are described in section 4.1.6 and
used at the end of the simulation model, so that the last order writes all the required data to the
Microsoft Excel® spreadsheet.
44
Simulation Model Construction
The simulation model construction can be very time consuming, depending on its detail and
complexity. The best approach is to start with a simple model and increasing the detail and
complexity as system knowledge advance. Verification and validation of the
methods used are discussed in this section.
Model Activity Areas
The simulation model construction is divided into several activity areas to better understand and
illustrate the logic of the model. The activity areas are as follows:
formation generation
and Output transfer
Order Information Generation
entities are generated, the first is created to read all the specified
ables and the second entity is specified as an order from a customer.
The first entity is created at time zero and runs through the read modules (Figure
variables as in the picking processing time of the different business units,
processing time and the total number of orders. Section 5.1.3 describes the dif
from the Microsoft Excel® spreadsheets. The variables are then incorporated
into the different distributions for the picking and packing processing times, these distributions
and 4.1.7. The variable that indicates the total number of orders is
the simulation model, so that the last order writes all the required data to the
The simulation model construction can be very time consuming, depending on its detail and
complexity. The best approach is to start with a simple model and increasing the detail and
. Verification and validation of the processes and
The simulation model construction is divided into several activity areas to better understand and
are generated, the first is created to read all the specified
Figure 15); it reads
nt business units, packing
describes the different variables
from the Microsoft Excel® spreadsheets. The variables are then incorporated
into the different distributions for the picking and packing processing times, these distributions
The variable that indicates the total number of orders is
the simulation model, so that the last order writes all the required data to the
Figure 15: Read Simulation Model Variables
The second entity is described as an
enters the first assign module (Figure
The entity then enters a separate module
module where the variable (Input_Taller) is added with a value of one (Input_Taller + 1).
duplicated entity reads (using read module)
value is either a 0 or a 1. If it is a 0 then it means that there are more orders
send from the pick pool (it is not he last order) and if it is a 1 then it is the last order.
decide module verify the value of the number
entity on (the statement is true);
pool. If it is not the last entity then it enters the second decide module.
the pick pool to picking in batches that are specified by the user (e.g.
time to picking). The second decide model
equal to that value which is specified by the user
equal then the entity follows the path explained above
(Input_Taller) is increased with 1.
If the second decide module is true (
module. The scan module scans the queue length at picking (Transport between stations
queues). It verifies if the value of the
to 2, if not then the scan module waits until the queue is smaller or equal to 2. If that occurs then
the entity is released by the scan module. The entity then enters the first assign module, where
the variable (Input_Taller) gets assigned the value zero. The same pr
until the last order is send on by the first decide module.
presentation.
45
: Read Simulation Model Variables
The second entity is described as an order made by a customer and is created at ti
Figure 16) where “zero” is assigned to a variable (Input_Taller).
separate module and is duplicated. It then enters the second assign
module where the variable (Input_Taller) is added with a value of one (Input_Taller + 1).
read module) from a Microsoft Excel® spreadsheet a value,
is either a 0 or a 1. If it is a 0 then it means that there are more orders
send from the pick pool (it is not he last order) and if it is a 1 then it is the last order.
the value of the number (0 or 1) and if that number is 1 then it sends the
; no more entities are duplicated as it is the last order from pick
If it is not the last entity then it enters the second decide module. The order
in batches that are specified by the user (e.g. 100 orders are sent
decide model verify the variable (Input_Taller),
specified by the user, the batch size value (e.g.
ollows the path explained above and the value of
increased with 1. (Input_Taller gets bigger until it equals the batch size value)
If the second decide module is true (equal to the user specified value) then it enters a scan
module. The scan module scans the queue length at picking (Transport between stations
value of the sum of the queue length divided by four is smaller or equal
then the scan module waits until the queue is smaller or equal to 2. If that occurs then
the entity is released by the scan module. The entity then enters the first assign module, where
the variable (Input_Taller) gets assigned the value zero. The same process occur as explained
until the last order is send on by the first decide module. See Figure 16
and is created at time zero. It
a variable (Input_Taller).
enters the second assign
module where the variable (Input_Taller) is added with a value of one (Input_Taller + 1). The
from a Microsoft Excel® spreadsheet a value, this
is either a 0 or a 1. If it is a 0 then it means that there are more orders that needs to be
send from the pick pool (it is not he last order) and if it is a 1 then it is the last order. The first
and if that number is 1 then it sends the
as it is the last order from pick
The orders are sent from
100 orders are sent at a
if the variable is
(e.g. 100). If it is not
and the value of the variable
(Input_Taller gets bigger until it equals the batch size value).
equal to the user specified value) then it enters a scan
module. The scan module scans the queue length at picking (Transport between stations
divided by four is smaller or equal
then the scan module waits until the queue is smaller or equal to 2. If that occurs then
the entity is released by the scan module. The entity then enters the first assign module, where
ocess occur as explained
16 for a graphical
Create single
entity
Figure 16: Duplication of Entity to the Number of Orders
The entity then enters the process of
Excel® spreadsheets. These attributes are order number, quantity and type of business
etc. description of all the input data is described in section
processing sequence, time and
read model. Figure 17 illustrates the path of the entity in obtaining the different attributes of
order. As can be seen in Figure
module is used to assign the current
analyse the simulation time. It gives the entity a time stamp of when it entered the simulation
model.
46
Last Order
: Duplication of Entity to the Number of Orders
The entity then enters the process of getting assigning different attributes from the Microsoft
Excel® spreadsheets. These attributes are order number, quantity and type of business
etc. description of all the input data is described in section 4.1. These attributes will dic
quantity. The attributes are read from the spreadsheet with a
illustrates the path of the entity in obtaining the different attributes of
Figure 17, when the entity enters the simulation model
module is used to assign the current simulation time to that entity, this attribute is later used to
analyse the simulation time. It gives the entity a time stamp of when it entered the simulation
Route to read
in attributes
from the Microsoft
Excel® spreadsheets. These attributes are order number, quantity and type of business units’
. These attributes will dictate
preadsheet with a
illustrates the path of the entity in obtaining the different attributes of each
, when the entity enters the simulation model an assign
attribute is later used to
analyse the simulation time. It gives the entity a time stamp of when it entered the simulation
Route from
Entity creation
Figure 17: Assigning and Reading Attributes to Entity
Appendix 10.4.1 illustrates the complete Arena® model for the order information generation
activity.
6.1.2 Pick Face
The entity assigned with different attributes also known as an order
all the different floors of the pick
and the order is send to all four. A decide module checks the attributed value of the transport
time on that level, if the number is greater
business units occur on that level, thus the order is send to that level
time for that level is zero then it means that no transporting of the business unit
level and the order is send to the dispose module.
47
: Assigning and Reading Attributes to Entity
illustrates the complete Arena® model for the order information generation
The entity assigned with different attributes also known as an order, gets duplicated
all the different floors of the pick face area. There are four levels of picking as stated previously
and the order is send to all four. A decide module checks the attributed value of the transport
time on that level, if the number is greater than zero then it means that
on that level, thus the order is send to that level for picking
time for that level is zero then it means that no transporting of the business unit
nd the order is send to the dispose module.
Route to Pick
phase
illustrates the complete Arena® model for the order information generation
gets duplicated and sent to
area. There are four levels of picking as stated previously
and the order is send to all four. A decide module checks the attributed value of the transport
than zero then it means that transportation of
for picking. If the transport
time for that level is zero then it means that no transporting of the business unit occurs on that
Route from
Reading
Attributes
Figure 18: Assigning Orders to different levels for picking
After defining that picking should take place on a level by the decide module, the order is send
on for processing of the transport
business units.
The order passes an assign node
as an attribute, this time stamp indicates the time the picking process for an order starts. At the
end of the picking process another assign module assigns the simulation time to that same
order, thus the beginning and ending times of picking is given to the order
write module writes the specified attributes to a Microsoft Excel® spread
analysis of the times an order spends in picking.
48
Route to Pick
Units
Route to Levels
B, C and D
to different levels for picking
After defining that picking should take place on a level by the decide module, the order is send
on for processing of the transport of the picker from each location and the actual picking of the
The order passes an assign node (Figure 19) where the simulation time is stamped on the order
as an attribute, this time stamp indicates the time the picking process for an order starts. At the
d of the picking process another assign module assigns the simulation time to that same
order, thus the beginning and ending times of picking is given to the order as
write module writes the specified attributes to a Microsoft Excel® spreadsheet for further
analysis of the times an order spends in picking.
Route to Pick
Units
After defining that picking should take place on a level by the decide module, the order is send
and the actual picking of the
where the simulation time is stamped on the order
as an attribute, this time stamp indicates the time the picking process for an order starts. At the
d of the picking process another assign module assigns the simulation time to that same
as attributes. The
sheet for further
Figure 19: Picking and Conveying Processes
The order enters a process model called transport between stations LA (or LB, LC, LD) where a
resource namely a picker is seized and delayed for a specific
the time the picker takes to walk from each location to the next to p
specific order. The time is calculated in Microsoft Excel® see section
calculation of the transport time of a picker for an order.
After the transportation process, the order then gets delayed from the same resource (picker).
The second delay is the process time for the picking of the business units
pick one unit). The quantity of business units is multiplied
specific business unit. The processing time (picking time) for footwear differs from that of the
other business unit, they are added together and form the picking processing time
processing times is described in secti
picked units are placed in totes (
placed on the conveyor and the picker resource is released in order to seize another order.
The conveying of the totes from picking to packing is at a constant velocity, a delay module is
used (Figure 19). The time it takes to convey the totes is directly dependant on the location
where the tote is placed by the picker on the conveyor. The calculation of the distance is
described in section 4.1.5 and the distance the conveyor must travel for that order
49
: Picking and Conveying Processes
a process model called transport between stations LA (or LB, LC, LD) where a
resource namely a picker is seized and delayed for a specific period of time. The delay time is
the time the picker takes to walk from each location to the next to pick business
. The time is calculated in Microsoft Excel® see section 4.1.2
calculation of the transport time of a picker for an order.
, the order then gets delayed from the same resource (picker).
The second delay is the process time for the picking of the business units (actual time it takes to
. The quantity of business units is multiplied with the processing time of that
specific business unit. The processing time (picking time) for footwear differs from that of the
, they are added together and form the picking processing time
processing times is described in section 4.1.3 refer to that section for more information.
picked units are placed in totes (container for transporting the ordered units), these totes are
the conveyor and the picker resource is released in order to seize another order.
The conveying of the totes from picking to packing is at a constant velocity, a delay module is
. The time it takes to convey the totes is directly dependant on the location
placed by the picker on the conveyor. The calculation of the distance is
the distance the conveyor must travel for that order
a process model called transport between stations LA (or LB, LC, LD) where a
time. The delay time is
business units for a
4.1.2 for details in
, the order then gets delayed from the same resource (picker).
(actual time it takes to
with the processing time of that
specific business unit. The processing time (picking time) for footwear differs from that of the
, they are added together and form the picking processing time. The
refer to that section for more information. The
), these totes are
the conveyor and the picker resource is released in order to seize another order.
The conveying of the totes from picking to packing is at a constant velocity, a delay module is
. The time it takes to convey the totes is directly dependant on the location
placed by the picker on the conveyor. The calculation of the distance is
the distance the conveyor must travel for that order is read into
the model through a read module (
multiplied with the constant velocity of the conveyor which is
time of the conveyor. Appendix
activity.
6.1.3 Pack Station
The tote that comes from the picking area via the conveyor passes an assign node where th
simulation time is given as an attribute to the tote. It is a time stamp of when the packing
process begins. The tote then enter a decide module, the module
specific packing station. The logic is that there are two packer
enters the packing area it must be assign to a pack
busy (if the number of resources busy are less then 2) then the tote is assign to that pack
station otherwise it checks the number of resources
that is less than two then it gets assign to pack station B, this is carried on until pack station J.
Pack station A will be the first to fill then B and so on.
Figure 20: Decide on packing station
50
the model through a read module (Figure 17). The distance which the conveyor
multiplied with the constant velocity of the conveyor which is 4.2sec/meter, giving the total delay
Appendix 10.4.2 illustrates the complete Arena® model for the pick
The tote that comes from the picking area via the conveyor passes an assign node where th
simulation time is given as an attribute to the tote. It is a time stamp of when the packing
process begins. The tote then enter a decide module, the module must assign the tote to a
specific packing station. The logic is that there are two packers at each packing station, if a tote
a it must be assign to a pack station. First the logic checks if packer A is
busy (if the number of resources busy are less then 2) then the tote is assign to that pack
he number of resources (packers) currently busy at pack station B
that is less than two then it gets assign to pack station B, this is carried on until pack station J.
Pack station A will be the first to fill then B and so on.
: Decide on packing station
conveyor must travel is
4.2sec/meter, giving the total delay
illustrates the complete Arena® model for the pick face
The tote that comes from the picking area via the conveyor passes an assign node where the
simulation time is given as an attribute to the tote. It is a time stamp of when the packing
assign the tote to a
at each packing station, if a tote
station. First the logic checks if packer A is
busy (if the number of resources busy are less then 2) then the tote is assign to that pack
at pack station B, if
that is less than two then it gets assign to pack station B, this is carried on until pack station J.
Figure 21 illustrates the actual packing of the units from the totes into boxes for shipment. The
tote enters the packing process and seizes a resources call a packer. The resource is seized,
delayed and released for a certain period of time. This pack processing time is calculated using
a mathematical formula see section
times and the distributions.
Figure 21: Packing Process
The boxes that have been filled with the units are conveyed to
conveyor uses to convey the boxes depend on the
packed. The further away the pack station is from the weigh station the longer the time it takes
the conveyor to convey the boxes.
the conveyor, because it moves at a constant
from the one station to the next.
the end of the packing process is stamped on the box as an attribute.
Appendix 10.4.3 illustrates the complete Arena® model for the pack
51
illustrates the actual packing of the units from the totes into boxes for shipment. The
tote enters the packing process and seizes a resources call a packer. The resource is seized,
nd released for a certain period of time. This pack processing time is calculated using
a mathematical formula see section 4.1.6 and 4.1.7 on the analysis of the pack processing
The boxes that have been filled with the units are conveyed to the weigh station. The time the
conveyor uses to convey the boxes depend on the location of pack station where the box was
The further away the pack station is from the weigh station the longer the time it takes
the conveyor to convey the boxes. This distance is calculated and multiplied with the
moves at a constant speed. The delay module is used as the conveyor
The boxes enter an assign module where the simulation time of
of the packing process is stamped on the box as an attribute.
illustrates the complete Arena® model for the pack station activity.
illustrates the actual packing of the units from the totes into boxes for shipment. The
tote enters the packing process and seizes a resources call a packer. The resource is seized,
nd released for a certain period of time. This pack processing time is calculated using
he pack processing
the weigh station. The time the
pack station where the box was
The further away the pack station is from the weigh station the longer the time it takes
his distance is calculated and multiplied with the speed of
The delay module is used as the conveyor
The boxes enter an assign module where the simulation time of
activity.
6.1.4 Despatch and Output transfer
The distance from the weigh stations to the despatching area is a fi
conveying times are constant and the time to weigh a box
time impediment a delay module is used
After the delay the boxes are batched together according to their order number, see section
4.1.4 on how many boxes must be batched per order.
assigned to the boxes; it places a time stamp on how long it took for an order to complete all the
processes. The different attributes are then
spreadsheet. These attributes are the time stamps that were placed on the orders to
time it takes to complete a specified process.
Figure 22: Delay, Batching and Output transfer processes
The box (order) enter a counter and
counter value is equal to the total number of order placed by the pick pool (
Max#Order). If it is true then it is the last order and all the resource utilisations are
through a write module to the Microsoft Excel® spreadsheet. The order is then despatched.
52
Output transfer
stations to the despatching area is a fixed length, thus the
conveying times are constant and the time to weigh a box is constant. Because of the constant
delay module is used to delay the boxes.
After the delay the boxes are batched together according to their order number, see section
on how many boxes must be batched per order. Next an attribute of the simulation time is
assigned to the boxes; it places a time stamp on how long it took for an order to complete all the
processes. The different attributes are then exported with a write module to a Microsoft Excel®
re the time stamps that were placed on the orders to
specified process.
: Delay, Batching and Output transfer processes
The box (order) enter a counter and after that a decide module. The decide module verify if the
counter value is equal to the total number of order placed by the pick pool (
Max#Order). If it is true then it is the last order and all the resource utilisations are
Microsoft Excel® spreadsheet. The order is then despatched.
xed length, thus the
Because of the constant
After the delay the boxes are batched together according to their order number, see section
ute of the simulation time is
assigned to the boxes; it places a time stamp on how long it took for an order to complete all the
with a write module to a Microsoft Excel®
re the time stamps that were placed on the orders to derive the
. The decide module verify if the
counter value is equal to the total number of order placed by the pick pool (the variable:
Max#Order). If it is true then it is the last order and all the resource utilisations are exported
Microsoft Excel® spreadsheet. The order is then despatched.
Route from
Conveyor
Figure 23: Despatching of orders
Appendix 10.4.4 illustrates the complete Arena® model for the despatch and output transfer
activity.
6.2 Verification and Validation of the Model
6.2.1 Verification
Verification is the process of ensuring that the mo
according to the modelling assumptions made.
is an important aspect in verifying the minor details of the simulation model. Debugging is a
major process that is involved in
that the model runs for the desired
6.2.1.1 Verification Techniques
A single entity is sent to enter the model, that entity is followed (step
to be sure that the model logic data is correct.
Animation is used as a visual representation of the activities performed by the model. The use
of various colours, graphs and assigning different picture to the status of each of the reso
(e.g. idle or busy), ensures that the entity flow can be followed.
53
orders
illustrates the complete Arena® model for the despatch and output transfer
Verification and Validation of the Model
Verification is the process of ensuring that the model behaves in the way it was intended,
according to the modelling assumptions made. A thorough investigation on the input parameters
is an important aspect in verifying the minor details of the simulation model. Debugging is a
in verification; all problems or bugs need to be cleared to ensure
that the model runs for the desired duration.
Verification Techniques
is sent to enter the model, that entity is followed (step-by-step) through the model
the model logic data is correct.
d as a visual representation of the activities performed by the model. The use
of various colours, graphs and assigning different picture to the status of each of the reso
the entity flow can be followed.
illustrates the complete Arena® model for the despatch and output transfer
del behaves in the way it was intended,
the input parameters
is an important aspect in verifying the minor details of the simulation model. Debugging is a
all problems or bugs need to be cleared to ensure
step) through the model
d as a visual representation of the activities performed by the model. The use
of various colours, graphs and assigning different picture to the status of each of the resources
Counters and Discrete-Change Variables
place by consulting the Output Analyz
through counters or variables.
SIMAN code verification is a method of validation. The
Processor (MOD) code can be examined for errors.
6.2.2 Validation
Validation is the task of ensuring that the model behaves the same as the
Validation can be described as the process of ensuring that the built simulation model is a good
representation of the physical syst
Validation does not mean that the model is exactly as the physical system. It does mean that
the model has been built sufficiently,
decisions concerning the physical system by testing them on the simulation model.
In order to validate a simulation model, the results from the model must be compared with the
results from the real system. If the results are similar then the validation is complete.
was done by comparing the outputs
54
Change Variables are also measures of entity flow. Verification takes
by consulting the Output Analyzer. These values should resemble the number of entities
is a method of validation. The Experiment Processor (EXP) and Model
Processor (MOD) code can be examined for errors.
Validation is the task of ensuring that the model behaves the same as the
can be described as the process of ensuring that the built simulation model is a good
of the physical system.
Validation does not mean that the model is exactly as the physical system. It does mean that
the model has been built sufficiently, representing, the physical system in order to make
decisions concerning the physical system by testing them on the simulation model.
In order to validate a simulation model, the results from the model must be compared with the
. If the results are similar then the validation is complete.
was done by comparing the outputs to the real system outputs.
are also measures of entity flow. Verification takes
er. These values should resemble the number of entities
Experiment Processor (EXP) and Model
Validation is the task of ensuring that the model behaves the same as the physical system.
can be described as the process of ensuring that the built simulation model is a good
Validation does not mean that the model is exactly as the physical system. It does mean that
representing, the physical system in order to make
decisions concerning the physical system by testing them on the simulation model.
In order to validate a simulation model, the results from the model must be compared with the
. If the results are similar then the validation is complete. Validation
6.3 Scenario Development
The main objective is to test the flexibility of the current system and to see if the demand
for the Soccer World Cup season can be met.
place on the system is also analysed.
scenarios to the simulation models itself in terms of logic mod
parameters imported from Microsoft Excel®.
kept roughly the same with minor changes throughout all the scenarios.
development is divided into three different
were developed.
6.3.1 Increased growth in forth
Information gained regarding growth in the
The purpose of the scenarios is to indicate the impact of growth of the quantity of units order on
the outbound process flow. The growth is
6.3.1.1 Scenario 1: A
System performance and input constraint
next three successive years. The base model is established.
6.3.1.2 Scenario 1: B
Increase the quantity of the units ordered by the customer
effective within three years, in 2012.
6.3.1.3 Scenario 1: C
Based on the changes made in scenario 2, increase the quantity of the units ordered by the
customers with another 5% (a
changes will be effective within three years, in 2015.
55
Scenario Development
The main objective is to test the flexibility of the current system and to see if the demand
for the Soccer World Cup season can be met. Future growth in demand and the pressure it will
is also analysed. There are only a few alterations made in the different
scenarios to the simulation models itself in terms of logic modifications, merely to the input
parameters imported from Microsoft Excel®. Model construction and operational activities are
the same with minor changes throughout all the scenarios.
development is divided into three different sections and from these sections different sce
forthcoming years
Information gained regarding growth in the quantity of units ordered was received from Nike.
The purpose of the scenarios is to indicate the impact of growth of the quantity of units order on
the outbound process flow. The growth is in 3 years incremental.
constraint remains the same as the current year
The base model is established.
Increase the quantity of the units ordered by the customers with 5%. These changes will be
effective within three years, in 2012.
de in scenario 2, increase the quantity of the units ordered by the
(a total increase of 10% from the current year 2009)
changes will be effective within three years, in 2015.
The main objective is to test the flexibility of the current system and to see if the demand levels
Future growth in demand and the pressure it will
There are only a few alterations made in the different
ifications, merely to the input
Model construction and operational activities are
the same with minor changes throughout all the scenarios. The scenario
different scenarios
received from Nike.
The purpose of the scenarios is to indicate the impact of growth of the quantity of units order on
as the current year (2009) for the
with 5%. These changes will be
de in scenario 2, increase the quantity of the units ordered by the
total increase of 10% from the current year 2009). These
6.3.2 Change in the number of resources
The resources (pickers and packers) are
closely correlated with the number of resources available for a specific process.
number of resources will decrease the resource utilisation but will increas
and this is of more importance to Nike.
compared, refer to section 3.2.
6.3.2.1 Scenario 2: A
The picker is the resource that is places the most strain in the model
pickers available. The picker resource capacity is increased by three to the to
picker of eleven.
6.3.2.2 Scenario 2: B
The picker resource capacity is increased further with four making the total capacity of the
picker resource to fifteen.
6.3.2.3 Scenario 2: C
The number of packer is underutilised
increase the resource utilisation. The number of pack stations is decreased from ten packing
stations to seven packing stations. This is an actual decrease of six packers because each pack
station has two packers assigned to it.
6.3.3 Change the number of
As explained in section 1.1.1 the
before picking tasks are assigned to each picker for processing
orders released from the pick pool is equal to 100
flexibility of the system. The flexibility of the system is an i
system will behave if numerous orders come through of high importance (Expedite orders).
These orders will resemble if a team of the Soccer World Cup goes through to the next round,
then an increase in sales are expected and must be delivered
56
Change in the number of resources
es (pickers and packers) are one of the key factors, the bottlenecks and queues are
closely correlated with the number of resources available for a specific process.
will decrease the resource utilisation but will increase the throughput time
of more importance to Nike. The Key Performance Indicators (KPI’s) will be
The picker is the resource that is places the most strain in the model, because of the number of
. The picker resource capacity is increased by three to the to
The picker resource capacity is increased further with four making the total capacity of the
underutilised thus a decrease in the number of pack
increase the resource utilisation. The number of pack stations is decreased from ten packing
stations to seven packing stations. This is an actual decrease of six packers because each pack
two packers assigned to it.
the number of orders released from the pick pool
the orders placed by customers are consolidated in a
are assigned to each picker for processing. The batch size of the number of
orders released from the pick pool is equal to 100. Decreasing this value will increase the
flexibility of the system. The flexibility of the system is an important factor, thus te
system will behave if numerous orders come through of high importance (Expedite orders).
These orders will resemble if a team of the Soccer World Cup goes through to the next round,
then an increase in sales are expected and must be delivered within the next day.
, the bottlenecks and queues are
closely correlated with the number of resources available for a specific process. Increasing the
e the throughput time
The Key Performance Indicators (KPI’s) will be
, because of the number of
. The picker resource capacity is increased by three to the total number of
The picker resource capacity is increased further with four making the total capacity of the
pack stations will
increase the resource utilisation. The number of pack stations is decreased from ten packing
stations to seven packing stations. This is an actual decrease of six packers because each pack
consolidated in a pick pool
The batch size of the number of
. Decreasing this value will increase the
mportant factor, thus testing how the
system will behave if numerous orders come through of high importance (Expedite orders).
These orders will resemble if a team of the Soccer World Cup goes through to the next round,
within the next day. The batch
size is changed from 100 order to 50 order
expedite orders can be handled more efficiently.
6.4 Possible follow-up S
6.4.1 Business unit locations
The distance a picker must walk in the picking station is relative to the
that must be picked. Fast moving items are a
conveyor on the ground floor, level A
level is a way to reduce travel time.
patterns will have a positive effect on the throughput tim
6.4.2 Conveyor Speed
The conveyor takes a very long time in conveying the units
pack station and then to dispatching. Thus analysis of the time the conveyor takes can be
measured and a trade-off study developed. The trade
throughput time of faster conveyors to the capital cost of new conveyors.
57
is changed from 100 order to 50 orders, thus increasing the flexibility of the system and
expedite orders can be handled more efficiently.
up Scenarios
Business unit locations
The distance a picker must walk in the picking station is relative to the classification
Fast moving items are assigned to locations as close as possible to the
conveyor on the ground floor, level A. Different patterns in the way the units are stored on each
level is a way to reduce travel time. Calculating the different travel times for the different
patterns will have a positive effect on the throughput time. Thus, optimising pick directives.
takes a very long time in conveying the units from pick stations to their assigned
station and then to dispatching. Thus analysis of the time the conveyor takes can be
off study developed. The trade-off study is between the incre
throughput time of faster conveyors to the capital cost of new conveyors.
, thus increasing the flexibility of the system and
classification of the items
s possible to the
the way the units are stored on each
Calculating the different travel times for the different
Thus, optimising pick directives.
ations to their assigned
station and then to dispatching. Thus analysis of the time the conveyor takes can be
off study is between the increases in the
7 Simulation Results
All the scenario runs consisted of one replication for the duration of completion of all the orders.
For information regarding the input data refer to section
section 6.3 and will be evaluated in the following section
from reports created by Arena® and
containing a statistical summary of specified data collection are generated after each simulation
run. Additional output date is shown in section Appendix
7.1 Increase growth in forthcoming years
The aim is to evaluate yearly throughput time
order to make correct assumptions.
7.1.1 Scenario 1: A
All constraints remain the same, thus the current system results is obtained in
the current output of Nike. It is also referred to t
Table 9: Base model results
Statistics
Throughput
Time
Time in
Process
58
Simulation Results and Recommendations
All the scenario runs consisted of one replication for the duration of completion of all the orders.
the input data refer to section 4. The scenarios are described in
and will be evaluated in the following section regarding output parameters attained
from reports created by Arena® and data exported to Microsoft Excel® spreadsheets.
containing a statistical summary of specified data collection are generated after each simulation
dditional output date is shown in section Appendix 10.3.
Increase growth in forthcoming years
The aim is to evaluate yearly throughput time of the orders and utilisation in each section in
order to make correct assumptions.
All constraints remain the same, thus the current system results is obtained in
lso referred to the base model.
Statistics Output Value (min)
Throughput Average 89.55
Average 79.03
Picking LA 90.65
Picking LB 77.62
Picking LC 54.34
Picking LD 60.3
Packing 6.16
All the scenario runs consisted of one replication for the duration of completion of all the orders.
rios are described in
output parameters attained
spreadsheets. Reports
containing a statistical summary of specified data collection are generated after each simulation
and utilisation in each section in
All constraints remain the same, thus the current system results is obtained in Table 9. It states
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Packer
A
Packer
B
Packer
C
58.3%
48.9%
37.6%
Av
era
ge
Resource Utilisation
Figure 24: Resource utilisation of base model
The picker is the resource that is utilised the most
(bottle neck).
7.1.2 Scenario 1: B
As expected the increase in the
throughput time and also the time in process.
59
Packer Packer
D
Packer
E
Packer
F
Packer
G
Packer
H
Packer
I
Packer
J
Picker
37.6%
23.8%
13.8%
7.1%2.2% 0.8% 0.2% 0.0%
77.0%
Resource Utilisation
: Resource utilisation of base model with 8 pickers
The picker is the resource that is utilised the most, indicating that it is the constraint resource
As expected the increase in the number of units ordered by the customers increased the
throughput time and also the time in process.
Picker
77.0%
, indicating that it is the constraint resource
number of units ordered by the customers increased the
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Packer
A
Packer
B
Packer
C
60.2%
48.5%
34.6%
Av
era
ge
Resource Utilisation
Table 10: Three year projection in quantity
Statistics
Throughput
Time
Time in
Process
The utilisation of the resources increased, with
5% increase of the number of units ordered by the customers.
Figure 25: Resource utilisation of three
60
Packer
D
Packer
E
Packer
F
Packer
G
Packer
H
Packer I Packer J
21.0%
11.8%
5.8%2.1% 0.8% 0.0% 0.0%
Resource Utilisation
: Three year projection in quantity of units ordered results
Statistics Output Value (min)
Throughput Average 95.12
Average 84.41
Picking LA 64.16
Picking LB 57.93
Picking LC 82.52
Picking LD 97.07
Packing 6.13
the resources increased, with more work that had to be done because of
units ordered by the customers.
: Resource utilisation of three year projection
Packer J Picker
0.0%
80.0%
re work that had to be done because of the
It is suggested that over the period of three years with an increase of 5% on the number of units
ordered by the customer, that the number of the pickers must be increased.
can be done on the number of picker resources required in order
throughput time. It is not necessary to increase the number of packer because the last four
packing stations utilisation’s is almost zero.
7.1.3 Scenario 1: C
As in the previous scenario the
increase of the number of units ordered from customers.
Table 11: Six year projection in quantity of units ordered
Statistics
Throughput
Time
Time in
Process
An increase in volume results in higher resource utilisation.
61
It is suggested that over the period of three years with an increase of 5% on the number of units
ordered by the customer, that the number of the pickers must be increased. Additional studies
can be done on the number of picker resources required in order to maintain the current
t is not necessary to increase the number of packer because the last four
packing stations utilisation’s is almost zero.
the previous scenario the throughput time and the time in process incre
increase of the number of units ordered from customers.
: Six year projection in quantity of units ordered results
Statistics Output Value (min)
Throughput Average 100.15
Average 89.1
Picking LA 102.65
Picking LB 86.13
Picking LC 62.37
Picking LD 67.49
Packing 6.15
An increase in volume results in higher resource utilisation.
It is suggested that over the period of three years with an increase of 5% on the number of units
Additional studies
to maintain the current
t is not necessary to increase the number of packer because the last four
time and the time in process increases with the
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Packer
A
Packer
B
Packer
C
64.4%
53.0%
36.5%
Av
era
ge
Resource Utilisation
Figure 26: Resource utilisation of six year projection
As the year pass more units are ordered thus more pickers must be employed to have the same
throughput time as the current year.
resources required in order to maintain the current throughput time.
62
Packer Packer
D
Packer
E
Packer
F
Packer
G
Packer
H
Packer
I
Packer
J
Picker
36.5%
21.7%
10.6%4.7%
1.5% 0.5% 0.1% 0.0%
86.2%
Resource Utilisation
: Resource utilisation of six year projection
year pass more units are ordered thus more pickers must be employed to have the same
throughput time as the current year. Additional studies can be done on the number of picker
resources required in order to maintain the current throughput time.
Picker
86.2%
year pass more units are ordered thus more pickers must be employed to have the same
Additional studies can be done on the number of picker
7.2 Change in resource capacity
The aim is to evaluate the throughput time of the orders and utilisation with the change in
resources. Refer to section 7.1.1
7.2.1 Scenario 2: A
The increase in the number of picker resources greatly reduced the throughput time. This is an
indication of the sensitivity of the sy
in all the times is illustrated.
Table 12: Increase picker resource capacity
Statistics
Throughput
Time
Time in
Process
The resource utilisation for the picker shows a
utilisation of the packers (except for packer A). This change in the utilisation is as predicted,
more picker are available thus more units
the packers work harder because
more important for Nike to decrease the throughput time
utilisations. In doing this more units can be sold and in turn more revenue.
63
resource capacity
The aim is to evaluate the throughput time of the orders and utilisation with the change in
7.1.1 for the base case of the simulation model.
The increase in the number of picker resources greatly reduced the throughput time. This is an
indication of the sensitivity of the system related to the number of picker resources. A reduction
: Increase picker resource capacity by three results
Statistics Output Value (min)
Throughput Average 72.78
Average 62.94
Picking LA 70
Picking LB 63.49
Picking LC 44.93
Picking LD 52.31
Packing 6.03
for the picker shows a decrease, but an increase is shown in the
utilisation of the packers (except for packer A). This change in the utilisation is as predicted,
more picker are available thus more units are pushed through the system. This in turn makes
s work harder because the units are sent from picking at a much higher rate.
more important for Nike to decrease the throughput time then to increase all the resource
. In doing this more units can be sold and in turn more revenue.
The aim is to evaluate the throughput time of the orders and utilisation with the change in
The increase in the number of picker resources greatly reduced the throughput time. This is an
stem related to the number of picker resources. A reduction
but an increase is shown in the
utilisation of the packers (except for packer A). This change in the utilisation is as predicted,
his in turn makes
the units are sent from picking at a much higher rate. It is
then to increase all the resource
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Packer
A
Packer
B
Packer
C
55.0%
49.2%
41.6%
Av
era
ge
Resource Utilisation
Figure 27: Resource utilisation with increase of three pickers
7.2.2 Scenario 2: B
Increasing the picker also indicated a reduction
The results indicate a drop in picker
increase in the packer utilisation indicates that more units are pushed through the system.
64
Packer Packer
D
Packer
E
Packer
F
Packer
G
Packer
H
Packer
I
Packer
J
Picker
41.6%
32.1%
22.1%
14.6%
8.8%4.7%
2.6% 1.4%
68.3%
Resource Utilisation
: Resource utilisation with increase of three pickers
also indicated a reduction in the throughput time and the time in process.
results indicate a drop in picker utilisation compared to pervious scenarios
increase in the packer utilisation indicates that more units are pushed through the system.
Picker
68.3%
in the throughput time and the time in process.
scenarios, but a greater
increase in the packer utilisation indicates that more units are pushed through the system.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Packer
A
Packer
B
Packer
C
68.5%63.8%
57.7%
Av
era
ge
Table 13: Increase picker resource capacity by seven results
Statistics
Throughput
Time
Time in
Process
The results obtained behaved as the previous scenario.
decreased the throughput time.
utilisation except for the packer resource.
Figure 28: Resource utilisation with increase of seven pic
65
Packer Packer
D
Packer
E
Packer
F
Packer
G
Packer
H
Packer
I
Packer
J
57.7%
51.0%
41.3%
31.9%
23.4%
15.3%10.7%
13.5%
Resource Utilisation
: Increase picker resource capacity by seven results
Statistics Output Value (min)
Throughput Average 57.48
Average 48.88
Picking LA 52.79
Picking LB 50.89
Picking LC 35.68
Picking LD 43.4
Packing 6.15
The results obtained behaved as the previous scenario. Increasing the resource capacity
decreased the throughput time. The values obtained are lower in the times and increase in the
utilisation except for the packer resource.
: Resource utilisation with increase of seven pickers
Packer Picker
13.5%
69.8%
Increasing the resource capacity
and increase in the
It is suggested that Nike employ more picker. More analysis can be done on the
but increasing the number of pickers
an increase in the utilisation of the packers
7.2.3 Scenario 2: C
Decreasing the number of packing station by three did not have an adverse effect on the system
as predicted. This is because the base
stations is minimal.
Table 14: Decrease packer resource capacity by three results
Statistics
Throughput
Time
Time in
Process
66
Nike employ more picker. More analysis can be done on the
the number of pickers shows a considerable decrease in the throughput time and
of the packers.
Decreasing the number of packing station by three did not have an adverse effect on the system
as predicted. This is because the base case shows that the utilisation for the last packing
packer resource capacity by three results
Statistics Output Value (min)
Throughput Average 88.41
Average 78.06
Picking LA 89.05
Picking LB 76.59
Picking LC 54.65
Picking LD 60.63
Packing 6.26
Nike employ more picker. More analysis can be done on the exact number,
in the throughput time and
Decreasing the number of packing station by three did not have an adverse effect on the system
utilisation for the last packing
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Packer A Packer B
57.5%
49.1%
Av
era
ge
Resource Utilisation
Figure 29: Resource utilisation with decrease of three packing stations
This illustrates that the number of packing station
with very minimal effect on the throughput time and the
the packing station only increases their utilisation with a small percentage to absorb the work of
the packing stations that have been taken away
67
Packer C Packer D Packer E Packer F Packer G Picker
38.2%
24.5%
15.1%
7.7%4.9%
76.9%
Resource Utilisation
: Resource utilisation with decrease of three packing stations
This illustrates that the number of packing stations may be reduced with up to as many as three,
with very minimal effect on the throughput time and the utilisation of the resources. The rest of
the packing station only increases their utilisation with a small percentage to absorb the work of
the packing stations that have been taken away.
Picker
76.9%
ced with up to as many as three,
utilisation of the resources. The rest of
the packing station only increases their utilisation with a small percentage to absorb the work of
7.3 Change the batch size of
The aim is to evaluate the throughput time and
the number of orders send from the pick
batch sizes reduce the total system work in progress
Decreasing the number of orders from, the pick pool that is
flexibility of the system. Expedite orders are dealt with in a minimal amount of time, increasing
the customer satisfaction and meeting with the high
Figure 30: Decrease in orders send
Statistics
Throughput
Time
Time in
Process
The throughput time and the time in process are reduced with the reduction of the number of
orders from the pick pool and the utilisation increased considerably. The last number of packer
show a zero or very low utilisation, this indicates
68
batch size of orders send from the pick pool
m is to evaluate the throughput time and the utilisation of the orders with the change in
send from the pick pool; this is done to test the systems flexibility
batch sizes reduce the total system work in progress
e number of orders from, the pick pool that is released for picking, increases the
flexibility of the system. Expedite orders are dealt with in a minimal amount of time, increasing
the customer satisfaction and meeting with the high needed flexibility of the Soccer World Cup.
s send from pick pool results
Statistics Output Value (min)
Throughput Average 61.17
Average 52.76
Picking LA 55.99
Picking LB 55.56
Picking LC 39.88
Picking LD 48.26
Packing 6.12
The throughput time and the time in process are reduced with the reduction of the number of
orders from the pick pool and the utilisation increased considerably. The last number of packer
show a zero or very low utilisation, this indicates that less packing stations are required.
send from the pick pool
with the change in
flexibility. Smaller
picking, increases the
flexibility of the system. Expedite orders are dealt with in a minimal amount of time, increasing
the Soccer World Cup.
The throughput time and the time in process are reduced with the reduction of the number of
orders from the pick pool and the utilisation increased considerably. The last number of packers
less packing stations are required.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Packer
A
Packer
B
Packer
C
67.7%
55.8%
41.4%
Av
era
ge
Resource Utilisation
Figure 31: Resource utilisation with decrease in order
It is suggested that Nike reduce the
will increase the flexibility which
Cup.
As stated previously it is more important for Nike to decrease the throughput time than to
increase the utilisation of the resources. If more units are delivered to the customers more units
will be sold and so increasing the revenue of Nike.
The results obtained through the scenario development indicate that the simulation model is a
true representation of the Nike’s
output results as predicted. The simulation model may be used to test and analyse several
different scenarios.
69
Packer Packer
D
Packer
E
Packer
F
Packer
G
Packer
H
Packer
I
Packer
J
Picker
41.4%
27.7%
16.5%
8.4%3.6% 1.4% 0.0% 0.0%
87.1%
Resource Utilisation
: Resource utilisation with decrease in orders send from pick pool
reduce the batch size of the number of orders released for picking
ill increase the flexibility which is especially important for the up and coming Soccer World
As stated previously it is more important for Nike to decrease the throughput time than to
e the utilisation of the resources. If more units are delivered to the customers more units
will be sold and so increasing the revenue of Nike.
The results obtained through the scenario development indicate that the simulation model is a
’s outbound process flow. The input parameters influenced the
output results as predicted. The simulation model may be used to test and analyse several
Picker
87.1%
size of the number of orders released for picking. This
is especially important for the up and coming Soccer World
As stated previously it is more important for Nike to decrease the throughput time than to
e the utilisation of the resources. If more units are delivered to the customers more units
The results obtained through the scenario development indicate that the simulation model is a
outbound process flow. The input parameters influenced the
output results as predicted. The simulation model may be used to test and analyse several
8 Conclusion
This project not only satisfied the objectives of increas
the demand of the Soccer World Cup will be met. The managing and understanding of the
intricate processes of the Nike distribution facility will lead to higher service levels, satisfied
customers and an efficient workforce.
The most suitable industrial engineering tool to assists the management in controlling the
factors that affect Nike’s performance is the use of simulation modelling. The literature study
point out the Arena® simulation model is the most suitabl
The conceptual design combined with the stated input parameter and business decisions
ensure that all of the relevant matters are included in the model.
The unstable demand of the Soccer World Cup period is
influence the structure of the simulation model.
represents the real world and several scenarios are tested.
execution of the project can be use
resources deployment will save Nike labour cost due to high productivity and utilisation in
different department. The system changes will reduce process cycle time and order lead time,
which in turn will result in increased customer satisfaction. Increasing the flexibility indicated that
more goods could be delivered in the same period of time, thus increasing the number of sales.
Through successful identification of the bottlenecks in current
able to recognise and prioritise investment opportunities. By investing in the most appropriate
areas, Nike will not only improve their plant utilisation
chasing the root cause for the plant utilisation optimisation.
Overall the project is a huge success and will yield significant cost reduction after
implementation in February 2010. Preliminary indications are that the project will yield a
reduction on unit processing costs
re-usable internally and further studies can be conducted on outbound process flow and
planning processes.
70
This project not only satisfied the objectives of increasing plant flexibility, but also proved that
the demand of the Soccer World Cup will be met. The managing and understanding of the
intricate processes of the Nike distribution facility will lead to higher service levels, satisfied
workforce.
The most suitable industrial engineering tool to assists the management in controlling the
factors that affect Nike’s performance is the use of simulation modelling. The literature study
point out the Arena® simulation model is the most suitable method for completing the project.
The conceptual design combined with the stated input parameter and business decisions
ensure that all of the relevant matters are included in the model.
demand of the Soccer World Cup period is some of the considerations that
influence the structure of the simulation model. After verification and validation the model
represents the real world and several scenarios are tested. The scenarios generated during the
execution of the project can be used to address the initial problems identified. The change in the
resources deployment will save Nike labour cost due to high productivity and utilisation in
different department. The system changes will reduce process cycle time and order lead time,
n turn will result in increased customer satisfaction. Increasing the flexibility indicated that
more goods could be delivered in the same period of time, thus increasing the number of sales.
Through successful identification of the bottlenecks in current operational activities, Nike will be
able to recognise and prioritise investment opportunities. By investing in the most appropriate
areas, Nike will not only improve their plant utilisation and flexibility, but will also save money by
lant utilisation optimisation.
Overall the project is a huge success and will yield significant cost reduction after
implementation in February 2010. Preliminary indications are that the project will yield a
reduction on unit processing costs and increase throughput capacity. The simulation model is
usable internally and further studies can be conducted on outbound process flow and
ing plant flexibility, but also proved that
the demand of the Soccer World Cup will be met. The managing and understanding of the
intricate processes of the Nike distribution facility will lead to higher service levels, satisfied
The most suitable industrial engineering tool to assists the management in controlling the
factors that affect Nike’s performance is the use of simulation modelling. The literature study
e method for completing the project.
The conceptual design combined with the stated input parameter and business decisions
some of the considerations that
erification and validation the model
The scenarios generated during the
The change in the
resources deployment will save Nike labour cost due to high productivity and utilisation in
different department. The system changes will reduce process cycle time and order lead time,
n turn will result in increased customer satisfaction. Increasing the flexibility indicated that
more goods could be delivered in the same period of time, thus increasing the number of sales.
operational activities, Nike will be
able to recognise and prioritise investment opportunities. By investing in the most appropriate
, but will also save money by
Overall the project is a huge success and will yield significant cost reduction after
implementation in February 2010. Preliminary indications are that the project will yield a
The simulation model is
usable internally and further studies can be conducted on outbound process flow and
9 References
ACSL. [online] http://www.acslsim.
Advanced continuous simulation language. [online] sim.sagepub.com.innopac.up.ac.za/cgi/content/abstract/26/3/72
Arena® Professional Edition. [online] http://www.arenasimulation.com/products/professional_edition.asp
Bowersox, DJ & Marien, EJ & Helfericdistribution systems: Discussion of the computer model
http://www.acslsim.com/products/modeling.shtml
Buzacott, JA & Shanthikumar, JG 1993, Hall, New Jersey.
Chase, RB & Jacobs, FR & Aquilano, NJ 2006, advantage with global cases, 11
Harrington, HJ & Tumay, K 2000,
Kelton, WD & Sadowski, RP & Sturrock, DT 2007, New York.
Klingstam, P & Gullander, P 1999, engineering, Elsevier Science B.V., Sweden. [online] http://www.sciencedirect.com.innopac.up.ac.za/science
71
http://www.acslsim.com/products, accessed on 13May 2009
Advanced continuous simulation language. [online] http://o-sim.sagepub.com.innopac.up.ac.za/cgi/content/abstract/26/3/72, accessed on 13May 2009
Professional Edition. [online] http://www.arenasimulation.com/products/professional_edition.asp, accessed on 14May 2009
Bowersox, DJ & Marien, EJ & Helferich, OK, Dynamic simulation model and planning physical
distribution systems: Discussion of the computer model. [online] http://www.acslsim.com/products/modeling.shtml, accessed on 11May 2009
Buzacott, JA & Shanthikumar, JG 1993, Stochastic models of manufacturing systems
Chase, RB & Jacobs, FR & Aquilano, NJ 2006, Operations management for competitive
, 11th edn, McGraw-Hill, New York.
arrington, HJ & Tumay, K 2000, Simulation modeling methods, McGraw-Hill, New York.
Kelton, WD & Sadowski, RP & Sturrock, DT 2007, Simulation with Arena, 4th edn, McGraw
Klingstam, P & Gullander, P 1999, Overview of simulation tools for computer-aided production
, Elsevier Science B.V., Sweden. [online] http://www.sciencedirect.com.innopac.up.ac.za/science, accessed on 13May 2009
13May 2009
accessed on 14May 2009
Dynamic simulation model and planning physical
Stochastic models of manufacturing systems, Prentice-
Operations management for competitive
Hill, New York.
edn, McGraw-Hill,
aided production
accessed on 13May 2009
Mehrotra, V., Fama, J., 2003.
opportunities. Proceedings of the 2003 Winter Simulation Conference. Edited by: Chick,
S.,Sanchez,P.J.,Ferrin,D.,Morrice, D.J.
ProModel Optimisation Software. [online] http://www.promodel.com/products/promodel/features.asp
Wikipedia – Acsl. [online] http://en.wikipedia.org/wiki/advanced_continuous_simulation_language2009
Wikipedia - Arena Rockwell Software. [online] accessed on 14May 2009
Winston, WL 2004, Introduction to probability models, operations research,
Thomson Brooks/Cole, Canada.
72
Mehrotra, V., Fama, J., 2003. Call centre simulation modelling: methods, challenges and
Proceedings of the 2003 Winter Simulation Conference. Edited by: Chick,
S.,Sanchez,P.J.,Ferrin,D.,Morrice, D.J.
ProModel Optimisation Software. [online] http://www.promodel.com/products/promodel/features.asp, accessed on 13May 2009
wiki/advanced_continuous_simulation_language, accessed on 11May
Arena Rockwell Software. [online] http://en.wikipedia.org/wiki/Rockwell_Arena
Introduction to probability models, operations research, vol.2,
Call centre simulation modelling: methods, challenges and
Proceedings of the 2003 Winter Simulation Conference. Edited by: Chick,
accessed on 13May 2009
accessed on 11May
http://en.wikipedia.org/wiki/Rockwell_Arena,
vol.2, 4th edn,
Sum
of T
rans
port
Tim
eCo
lum
n La
bels
Row
Lab
els
1-1
1-2
1-3
1-4
1-5
1-6
1-7
1-8
1-9
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
3-9
Gra
nd T
otal
1-1
015
3045
6075
9010
512
011
2641
5671
8610
111
613
122
3752
6782
9711
212
714
219
17
1-2
150
1530
4560
7590
105
2611
2641
5671
8610
111
637
2237
5267
8297
112
127
1602
1-3
3015
015
3045
6075
9041
2611
2641
5671
8610
152
3722
3752
6782
9711
213
77
1-4
4530
150
1530
4560
7556
4126
1126
4156
7186
6752
3722
3752
6782
9712
42
1-5
6045
3015
015
3045
6071
5641
2611
2641
5671
8267
5237
2237
5267
8211
97
1-6
7560
4530
150
1530
4586
7156
4126
1126
4156
9782
6752
3722
3752
6712
42
1-7
9075
6045
3015
015
3010
186
7156
4126
1126
4111
297
8267
5237
2237
5213
77
1-8
105
9075
6045
3015
015
116
101
8671
5641
2611
2612
711
297
8267
5237
2237
1602
10 Appendices
10.1 Input Data Analysis
Table 15: Picker Transport Time Table for Level A
73
1-8
105
9075
6045
3015
015
116
101
8671
5641
2611
2612
711
297
8267
5237
2237
1602
1-9
120
105
9075
6045
3015
013
111
610
186
7156
4126
1114
212
711
297
8267
5237
2219
17
2-1
1126
4156
7186
101
116
131
015
3045
6075
9010
512
011
2641
5671
8610
111
613
118
18
2-2
2611
2641
5671
8610
111
615
015
3045
6075
9010
526
1126
4156
7186
101
116
1503
2-3
4126
1126
4156
7186
101
3015
015
3045
6075
9041
2611
2641
5671
8610
112
78
2-4
5641
2611
2641
5671
8645
3015
015
3045
6075
5641
2611
2641
5671
8611
43
2-5
7156
4126
1126
4156
7160
4530
150
1530
4560
7156
4126
1126
4156
7110
98
2-6
8671
5641
2611
2641
5675
6045
3015
015
3045
8671
5641
2611
2641
5611
43
2-7
101
8671
5641
2611
2641
9075
6045
3015
015
3010
186
7156
4126
1126
4112
78
2-8
116
101
8671
5641
2611
2610
590
7560
4530
150
1511
610
186
7156
4126
1126
1503
2-9
131
116
101
8671
5641
2611
120
105
9075
6045
3015
013
111
610
186
7156
4126
1118
18
3-1
2237
5267
8297
112
127
142
1126
4156
7186
101
116
131
015
3045
6075
9010
512
019
17
3-2
3722
3752
6782
9711
212
726
1126
4156
7186
101
116
150
1530
4560
7590
105
1602
3-3
5237
2237
5267
8297
112
4126
1126
4156
7186
101
3015
015
3045
6075
9013
77
3-4
6752
3722
3752
6782
9756
4126
1126
4156
7186
4530
150
1530
4560
7512
42
3-5
8267
5237
2237
5267
8271
5641
2611
2641
5671
6045
3015
015
3045
6011
97
3-6
9782
6752
3722
3752
6786
7156
4126
1126
4156
7560
4530
150
1530
4512
42
3-7
112
9782
6752
3722
3752
101
8671
5641
2611
2641
9075
6045
3015
015
3013
77
3-8
127
112
9782
6752
3722
3711
610
186
7156
4126
1126
105
9075
6045
3015
015
1602
3-9
142
127
112
9782
6752
3722
131
116
101
8671
5641
2611
120
105
9075
6045
3015
019
17
Gra
nd T
otal
1917
1602
1377
1242
1197
1242
1377
1602
1917
1818
1503
1278
1143
1098
1143
1278
1503
1818
1917
1602
1377
1242
1197
1242
1377
1602
1917
3952
8
Input Data Analysis
: Picker Transport Time Table for Level A
Gra
nd T
otal
1917
1602
1377
1242
1197
1242
1377
1602
1917
1818
1503
1278
1143
1098
1143
1278
1503
1818
1917
1602
1377
1242
1197
1242
1377
1602
1917
3952
8
Su
m o
f T
ran
spo
rtT
ime
Co
lum
n L
ab
els
Ro
w L
ab
els
1-1
1-1
01
-11
1-1
21
-21
-31
-41
-51
-61
-71
-81
-92
-12
-10
2-1
12
-12
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
3-1
3-1
03
-11
3-1
23
-23
-33
-43
-53
-63
-73
-83
-9G
ran
d T
ota
l
1-1
01
17
13
01
43
13
26
39
52
65
78
91
10
41
41
31
14
41
57
27
40
53
66
79
92
10
51
18
28
14
51
58
17
14
15
46
78
09
31
06
11
91
32
30
78
1-1
01
17
01
32
61
04
91
78
65
52
39
26
13
13
11
42
74
01
18
10
59
27
96
65
34
02
71
45
28
41
54
13
21
19
10
69
38
06
75
44
12
37
6
1-1
11
30
13
01
31
17
10
49
17
86
55
23
92
61
44
27
14
27
13
11
18
10
59
27
96
65
34
01
58
41
28
41
14
51
32
11
91
06
93
80
67
54
26
88
1-1
21
43
26
13
01
30
11
71
04
91
78
65
52
39
15
74
02
71
41
44
13
11
18
10
59
27
96
65
31
71
54
41
28
15
81
45
13
21
19
10
69
38
06
73
07
8
1-2
13
10
41
17
13
00
13
26
39
52
65
78
91
27
11
81
31
14
41
42
74
05
36
67
99
21
05
41
13
21
45
15
82
84
15
46
78
09
31
06
11
92
68
8
1-3
26
91
10
41
17
13
01
32
63
95
26
57
84
01
05
11
81
31
27
14
27
40
53
66
79
92
54
11
91
32
14
54
12
84
15
46
78
09
31
06
23
76
1-4
39
78
91
10
42
61
30
13
26
39
52
65
53
92
10
51
18
40
27
14
27
40
53
66
79
67
10
61
19
13
25
44
12
84
15
46
78
09
32
14
2
Table 16: Picker Transport Time Table for Level B, C, D
74
1-4
39
78
91
10
42
61
30
13
26
39
52
65
53
92
10
51
18
40
27
14
27
40
53
66
79
67
10
61
19
13
25
44
12
84
15
46
78
09
32
14
2
1-5
52
65
78
91
39
26
13
01
32
63
95
26
67
99
21
05
53
40
27
14
27
40
53
66
80
93
10
61
19
67
54
41
28
41
54
67
80
19
86
1-6
65
52
65
78
52
39
26
13
01
32
63
97
96
67
99
26
65
34
02
71
42
74
05
39
38
09
31
06
80
67
54
41
28
41
54
67
19
08
1-7
78
39
52
65
65
52
39
26
13
01
32
69
25
36
67
97
96
65
34
02
71
42
74
01
06
67
80
93
93
80
67
54
41
28
41
54
19
08
1-8
91
26
39
52
78
65
52
39
26
13
01
31
05
40
53
66
92
79
66
53
40
27
14
27
11
95
46
78
01
06
93
80
67
54
41
28
41
19
86
1-9
10
41
32
63
99
17
86
55
23
92
61
30
11
82
74
05
31
05
92
79
66
53
40
27
14
13
24
15
46
71
19
10
69
38
06
75
44
12
82
14
2
2-1
14
13
11
44
15
72
74
05
36
67
99
21
05
11
80
11
71
30
14
31
32
63
95
26
57
89
11
04
14
13
11
44
15
72
74
05
36
67
99
21
05
11
82
91
0
2-1
01
31
14
27
40
11
81
05
92
79
66
53
40
27
11
70
13
26
10
49
17
86
55
23
92
61
31
31
14
27
40
11
81
05
92
79
66
53
40
27
22
08
2-1
11
44
27
14
27
13
11
18
10
59
27
96
65
34
01
30
13
01
31
17
10
49
17
86
55
23
92
61
44
27
14
27
13
11
18
10
59
27
96
65
34
02
52
0
2-1
21
57
40
27
14
14
41
31
11
81
05
92
79
66
53
14
32
61
30
13
01
17
10
49
17
86
55
23
91
57
40
27
14
14
41
31
11
81
05
92
79
66
53
29
10
2-2
27
11
81
31
14
41
42
74
05
36
67
99
21
05
13
10
41
17
13
00
13
26
39
52
65
78
91
27
11
81
31
14
41
42
74
05
36
67
99
21
05
25
20
2-3
40
10
51
18
13
12
71
42
74
05
36
67
99
22
69
11
04
11
71
30
13
26
39
52
65
78
40
10
51
18
13
12
71
42
74
05
36
67
99
22
20
8
2-4
53
92
10
51
18
40
27
14
27
40
53
66
79
39
78
91
10
42
61
30
13
26
39
52
65
53
92
10
51
18
40
27
14
27
40
53
66
79
19
74
2-5
66
79
92
10
55
34
02
71
42
74
05
36
65
26
57
89
13
92
61
30
13
26
39
52
66
79
92
10
55
34
02
71
42
74
05
36
61
81
8
2-6
79
66
79
92
66
53
40
27
14
27
40
53
65
52
65
78
52
39
26
13
01
32
63
97
96
67
99
26
65
34
02
71
42
74
05
31
74
0
2-7
92
53
66
79
79
66
53
40
27
14
27
40
78
39
52
65
65
52
39
26
13
01
32
69
25
36
67
97
96
65
34
02
71
42
74
01
74
0
2-8
10
54
05
36
69
27
96
65
34
02
71
42
79
12
63
95
27
86
55
23
92
61
30
13
10
54
05
36
69
27
96
65
34
02
71
42
71
81
8
2-9
11
82
74
05
31
05
92
79
66
53
40
27
14
10
41
32
63
99
17
86
55
23
92
61
30
11
82
74
05
31
05
92
79
66
53
40
27
14
19
74
3-1
28
14
51
58
17
14
15
46
78
09
31
06
11
91
32
14
13
11
44
15
72
74
05
36
67
99
21
05
11
80
11
71
30
14
31
32
63
95
26
57
89
11
04
30
78
3-1
01
45
28
41
54
13
21
19
10
69
38
06
75
44
11
31
14
27
40
11
81
05
92
79
66
53
40
27
11
70
13
26
10
49
17
86
55
23
92
61
32
37
6
3-1
11
58
41
28
41
14
51
32
11
91
06
93
80
67
54
14
42
71
42
71
31
11
81
05
92
79
66
53
40
13
01
30
13
11
71
04
91
78
65
52
39
26
26
88
3-1
21
71
54
41
28
15
81
45
13
21
19
10
69
38
06
71
57
40
27
14
14
41
31
11
81
05
92
79
66
53
14
32
61
30
13
01
17
10
49
17
86
55
23
93
07
8
3-2
41
13
21
45
15
82
84
15
46
78
09
31
06
11
92
71
18
13
11
44
14
27
40
53
66
79
92
10
51
31
04
11
71
30
01
32
63
95
26
57
89
12
68
8
3-3
54
11
91
32
14
54
12
84
15
46
78
09
31
06
40
10
51
18
13
12
71
42
74
05
36
67
99
22
69
11
04
11
71
30
13
26
39
52
65
78
23
76
3-4
67
10
61
19
13
25
44
12
84
15
46
78
09
35
39
21
05
11
84
02
71
42
74
05
36
67
93
97
89
11
04
26
13
01
32
63
95
26
52
14
2
3-5
80
93
10
61
19
67
54
41
28
41
54
67
80
66
79
92
10
55
34
02
71
42
74
05
36
65
26
57
89
13
92
61
30
13
26
39
52
19
86
3-6
93
80
93
10
68
06
75
44
12
84
15
46
77
96
67
99
26
65
34
02
71
42
74
05
36
55
26
57
85
23
92
61
30
13
26
39
19
08
3-7
10
66
78
09
39
38
06
75
44
12
84
15
49
25
36
67
97
96
65
34
02
71
42
74
07
83
95
26
56
55
23
92
61
30
13
26
19
08
3-8
11
95
46
78
01
06
93
80
67
54
41
28
41
10
54
05
36
69
27
96
65
34
02
71
42
79
12
63
95
27
86
55
23
92
61
30
13
19
86
3-9
13
24
15
46
71
19
10
69
38
06
75
44
12
81
18
27
40
53
10
59
27
96
65
34
02
71
41
04
13
26
39
91
78
65
52
39
26
13
02
14
2
Gra
nd
To
tal
30
78
23
76
26
88
30
78
26
88
23
76
21
42
19
86
19
08
19
08
19
86
21
42
29
10
22
08
25
20
29
10
25
20
22
08
19
74
18
18
17
40
17
40
18
18
19
74
30
78
23
76
26
88
30
78
26
88
23
76
21
42
19
86
19
08
19
08
19
86
21
42
83
05
2
: Picker Transport Time Table for Level B, C, D
Order Number
1
2
3
4
5
6
7
8
9
10
11
12
13
Order Number LA_Other LA_Footware
1 0
2 13
3 0
4 0
5 24
6 0
7 0
8 0
9 1
10 19
11 0
12 0
13 0
14 0
15 0
Table 17: Extraction from Total Transport Time for Each Level in Picking
Table 18: Extraction from Quantity
75
A B C D Grand Total
0 0 0 75.5 75.5
33.5 0 0 0 33.5
93.5 0 0 0 93.5
59.5 0 0 0 59.5
59.5 0 0 0 59.5
0 0 74.5 59.5 134
0 63.5 0 0 63.5
0 75.5 0 0 75.5
0 49.5 0 0 49.5
63.5 0 75.5 0 139
0 101.5 0 0 101.5
0 0 49.5 0 49.5
37.5 0 0 0 37.5
LA_Footware LB_Other LB_Footware LC_Other LC_Footware LD_Other
0 0 0 0 0
0 0 12 0 0
12 0 0 0 24
15 0 0 0 0
0 0 0 0 0
0 0 0 134 0
0 11 0 0 0
0 0 30 0 0
0 2 0 0 9
0 0 0 0 0
0 24 0 0 0
0 0 0 0 43
3 0 0 0 0
1 0 0 0 0
0 1 0 108 0
Total Transport Time for Each Level in Picking
Quantity and Business Units per Order on each Level
LD_Other LD_Footware
7 0
0 0
0 0
0 0
0 0
0 0
0 0
0 5
0 0
0 0
0 0
0 0
0 0
0 0
0 0
Order Number A
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Order Number
Table 19: Extraction from Number to be batched on a Level
Table 20: Extraction Distance the Conveyor travel on a Level
76
B C D Batch
0 0 0 1 1
1 1 0 0 2
1 0 1 0 2
1 0 0 0 1
1 0 0 0 1
0 0 1 0 1
0 1 0 1 2
0 1 0 0 1
1 1 1 0 3
1 0 0 0 1
0 1 0 0 1
0 0 1 0 1
1 0 0 0 1
1 0 0 0 1
0 1 1 0 2
Order Number A B C D
1 0 0 0 61
2 30 0 0 0
3 54 0 0 0
4 36 0 0 0
5 36 0 0 0
6 0 0 66 0
7 0 43 0 0
8 0 61 0 0
9 0 52 0 0
10 18 0 0 0
11 0 70 0 0
12 0 0 52 0
13 24 0 0 0
14 36 0 0 0
15 0 29 0 0
Number to be batched on a Level
Distance the Conveyor travel on a Level
10.2 Probability Distributions
10.2.1 Picking Process Time
Distribution Summary
Distribution: Triangular
Expression: TRIA (1, 19, 48)
Square Error: 0.003419
Data Summary
Number of Data Points = 13
Min Data Value = 1
Max Data Value = 48
Sample Mean = 19
Sample Std Dev = 8.8
Histogram Summary
Histogram Range = 1 to
Number of Intervals = 11
77
Distributions
Picking Process Time for Business Unit “Other” Distribution
= 130
= 1
48
19
= 8.8
1 to 48
11
10.2.2 Picking Process Time for Business Unit
Distribution Summary
Distribution: Triangular
Expression: TRIA (3, 11, 26)
Square Error: 0.004735
Data Summary
Number of Data Points = 130
Min Data Value = 3
Max Data Value = 26
Sample Mean = 11
Sample Std Dev = 4.1
Histogram Summary
Histogram Range = 3
Number of Intervals = 11
78
Picking Process Time for Business Unit “Footwear” Distribution
= 130
= 3
= 26
= 11
= 4.1
= 3 to 26
= 11
Distribution
10.2.3 Packing Process Tim
Distribution Summary
Distribution: Normal
Expression: NORM (287, 98.2
Square Error: 0.005179
Data Summary
Number of Data Points = 150
Min Data Value = 0.1
Max Data Value = 542
Sample Mean = 287
Sample Std Dev = 98.2
Histogram Summary
Histogram Range = 0.1 to 542
Number of Intervals = 12
79
Packing Process Time Distribution
, 98.2)
= 150
= 0.1
= 542
= 287
= 98.2
= 0.1 to 542
= 12
10.3 Simulation Results
10.3.1 Queue average waiting time per entity
Waiting Time in Queue (Avg)
Increase Growth
Base
Model
Scenario
Batch Order same
OrderNumber 6.122 6.376
Packing Station A 0
Packing Station B 0
Packing Station C 0
Packing Station D 0
Packing Station E 0
Packing Station F 0
Packing Station G 0
Packing Station H 0
Packing Station I 0
Packing Station J 0
Transport
between Stations
LA
76.407 82.43
Transport
between Stations
LB
54.37 58.392
Transport
between Stations
LC
39.336 42.666
Transport
between Stations
LD
35.659 38.735
80
Queue average waiting time per entity
Waiting Time in Queue (Avg)
Increase Growth Change #Resources
Scenario
2
Scenario
3
Scenario
1
Scenario
2
Scenario
3
6.376 6.767 4.535 3.754 5.654
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 3.207
0 0 0 0 N/A
0 0 0 0 N/A
0 0 1.047 6.161 N/A
82.43 87.56 55.787 39.232 75.108
58.392 61.944 39.894 27.526 53.391
42.666 46.654 30.353 20.549 39.652
38.735 41.892 27.563 18.953 36.017
Change
#Units
send
Scenario
Scenario 1
5.654 3.76
0
0
0
0
0
0
3.207 0
N/A 0
N/A 0
N/A 0
75.108 41.974
53.391 32.145
39.652 25.499
36.017 24.058
10.3.2 Number of entities entering
Number in Process (Avg)
Increase Growth
Base
Model
Scenario
Packing Station A 841 917
Packing Station B 704 737
Packing Station C 548 533
Packing Station D 362 317
Packing Station E 202 176
Packing Station F 102
Packing Station G 34
Packing Station H 14
Packing Station I 3
Packing Station J 0
Picking Process LA 1430 1430
Picking Process LB 669 669
Picking Process LC 393 393
Picking Process LD 318 318
81
Number of entities entering a process
Increase Growth Change #Resources
Scenario
2
Scenario
3
Scenario
1
Scenario
2
Scenario
3
917 937 663 504 831
737 773 592 481 692
533 538 511 431 547
317 312 388 378 352
176 151 267 297 215
87 69 180 246 105
31 21 104 176 68
12 8 57 120 N/A
0 1 32 80 N/A
0 0 16 97 N/A
1430 1430 1430 1430 1430
669 669 669 669 669
393 393 393 393 393
318 318 318 318 318
Change #Units
send
Scenario Scenario 1
856
690
526
340
216
114
49
17
2
0
1430 1430
669
393
318
10.4 The Simulation Parts
10.4.1 Order information generation
82
The Simulation Parts
Order information generation phase
10.4.2 Pick Face
83
10.4.3 Pack Station
84
10.4.4 Despatch and Output transfer
85
Despatch and Output transfer phase